Datasets:

health360

/

Healix-Shot

like 4

events for this study. Myocardial Infarction Myocardial Infarction will be defined as either: 1. Q-wave MI: Development of new (i.e., not present on the subject's ECG before allocation) pathological Q-waves in 2 or more leads lasting ≥ 0.04 seconds with post procedure CK-MB levels elevated above normal. 2. Non-Q-Wave MI: De novo elevation of CK Total levels > 2.0 × ULN without the presence of new Qwaves (not present on the subject's ECG before allocation). If CK-MB performed, it must be positive. For subjects undergoing bypass surgery, a perioperative MI will be defined as (a) Total CK-MB > 5× upper limits of local laboratory normal, or (b) Presence of new pathologic Q waves (as defined above). Stent Thrombosis NOTE: Data will be collected which will allow for reporting per Boston Scientific's Historical (TAXUS IV and V) stent thrombosis definition as well as the Academic Research Consortium stent thrombosis definition [12]. Boston Scientific Historical (TAXUS IV and V [1,7]) Stent Thrombosis Definition: The occurrence of any of the following: 1. Clinical presentation of acute coronary syndrome with angiographic evidence of stent thrombosis: • Angiographic documentation of acute complete occlusion (TIMI flow 0 or 1) of the treated area in a previously successfully treated artery (TIMI flow 2 to 3 immediately after stent placement and diameter stenosis ≤ 30%) and/or • Angiographic documentation of a flow limiting thrombus within or adjacent to the successfully treated lesion. 2. Acute MI in the distribution of the treated vessel. 3. Cardiac death within the first 30 days post index procedure (without other obvious cause) is considered a surrogate for stent thrombosis when angiography is not available. Stent thrombosis will be classified as follows: 1. "Confirmed stent thrombosis" for the description of above events with angiographic evidence. 2. "Presumed stent thrombosis" for the description of above events in the absence of an angiography (i. e., such as in the case of death without autopsy). Academic Research Consortium Definition: Stent thrombosis will also be defined per the Definite, Probable, and Possible definitions described in Cutlip et al., 2007 [12]. Target Lesion Failure (TLF) Any ischemia-driven revascularization of the target lesion (TLR), MI (Q-wave and non-Q-wave) related to the target vessel, or (cardiac) death related to the target vessel. For the purposes of this protocol, if it cannot be determined with certainty whether MI or death was related to the target vessel, it will be considered TLF. Target Lesion Revascularization (TLR) Target Lesion Revascularization is defined as any ischemia-driven repeat percutaneous intervention (to improve blood flow) of the successfully treated target lesion or bypass surgery of the target vessel with a graft distally to the successfully treated target lesion. A target lesion revascularization will be considered as ischemia-driven if the target lesion diameter stenosis is ≥ 50% by QCA and there is presence of clinical or functional ischemia which cannot be explained by other coronary or graft lesions. Clinical or functional ischemia is any of the following: 1. The subject has a positive functional study corresponding to the area served by the target lesion 2. The subject has ischemic ECG changes at rest in a distribution consistent with the target vessel 3. The subject has ischemic symptoms referable to the target lesion. A target lesion revascularization will be considered as ischemia-driven if the lesion diameter stenosis is ≥ 70% by QCA even in the absence of clinical or functional ischemia. Target Vessel Failure (TVF) Any ischemia-driven revascularization of the target vessel, MI (Q-and non-Q-wave) related to the target vessel, or death related to the target vessel. For the purposes of this protocol, if it cannot be determined with certainty whether MI or death was related to the target vessel, it will be considered TVF. Target Vessel Revascularization (TVR) Presence of any Target Lesion Revascularization or Target Vessel Revascularization Remote. Target Vessel Revascularization, Remote (TVR, Non-TLR) Target Vessel Revascularization, Non-TLR is defined as any ischemia-driven repeat percutaneous intervention (to improve blood flow) or bypass surgery of not previously existing lesions ≥ 50% by QCA in the target vessel, excluding the target lesion. A target vessel revascularization will be considered ischemia-driven if the target vessel diameter stenosis is ≥ 50% by QCA and any of the following are present: 1. The subject has a positive functional study corresponding to the area served by the target vessel 2. The subject has ischemic ECG changes at rest in a distribution consistent with the target vessel 3. The subject has ischemic symptoms referable to the target vessel. A target vessel revascularization will also be considered as ischemia-driven if the lesion diameter stenosis is ≥ 70% even in the absence of clinical or functional ischemia. Technical Success Technical success is defined as successful delivery and deployment of the study stent to the target vessel, without balloon rupture or stent embolization. Identifying the impact of the confinement of Covid-19 on emotional-mood and behavioural dimensions in children and adolescents with attention deficit hyperactivity disorder (ADHD) The current study examined the impact of the lockdown due to the Covid-19 disease on mood state and behaviours of children and adolescents with ADHD. Nine hundred ninety-two parents of children and adolescents with ADHD filled out an anonymous online survey through the ADHD family association website. The survey investigated the degree of severity of six emotional and mood states (sadness, boredom, little enjoyment/interest, irritability, temper tantrums, anxiety) and five disrupted behaviours (verbal and physical aggression, argument, opposition, restlessness) based on their frequency/week (absent; low: 1–2 days/week; moderate: 3–4 days/week; severe: 5–7 days/week) before and during the lockdown. Important fluctuations were found in all dimensions during the lockdown independently by the severity degree. Subjects with previous low severity degree of these behaviors significantly worsened in almost all dimensions during the lockdown. On the contrary, ADHD patients with moderate and severe degree showed important improvement during the lockdown. Little enjoyment/interests and boredom resulted the dimensions more strongly affected by the condition of restriction, overall in children. Children vs. adolescents showed substantially similar trend but the former resulted significantly more vulnerable to emotive changes. The results provided both the individuation of domains affected, and the indirect benefits produced by restriction condition. Introduction The confinement due to the Covid-19 disease has represented a great challenge for children and adolescent of the entire world because it required a sudden adaptive change in daily lifestyle caused by imposed restrictions within familiar context. Italy is one of the major COVID-19 hotspots; to prevent disease spread, the lockdown started on March 9, 2020 until May 18 (seventy-one days) forcing people into home confinement and imposed restrictions on the movement of individuals in the entire national territory; but some restriction continued until June 15. The lockdown on the entire Italian territory was total and the population adhered very well to the confinement rules. The interruption of social relationships, a reduction of physical activities as well as the academic and normal working activities, has caused important disruptions of family routines, with alteration of circadian rhythm (Cellini et al., 2020) and eating habits (Pietrobelli et al., 2020) in all the countries affected by Covid-19. This unfavorable change has negatively affected mental health of the general population, especially of children and adolescents. Recent studies have shown an increase of irritability due to the prolonged boredom times in the general population with rapid swing of mood as well as behavioural problems. Cao et al., 2020). Children and adolescents with attention deficit hyperactivity disorders (ADHD) might be potentially vulnerable to the distress caused by lockdown due to Covid-19 and the European ADHD Guidelines Group alerted for ADHD management during the pandemic period, highlighting this investigation as a field of high priority. ADHD patients have intolerance for uncertainty, and they face difficulties in following instructions and understanding the complexity of the pandemic situation. Cortese et al., 2020, reported that the enforced condition at home and the unfriendly environment, altering their regular routine, could increase the chances of more severe hyperactivity and impulsive behaviors causing difficulty for the caregivers to engage these children in meaningful activities. Some studies confirmed the worsening of ADHD symptoms (Zhang et al., 2020), but, on the other hand, other authors reported improvements in restlessness and in the length of time of study in relation to a decrease of distress created by rhythm imposed of scholastic activities (Bobo et al., 2020). The authors stated that intra familiar environmental enforced condition could lead to a relaxation overall for children and adolescents with ADHD that experienced several external stressors. Studies performed on general population of children and adolescents showed that the pandemic and lockdown had a great impact on emotion and social relationships (Jiao et al., 2020;Lee, 2020). Nevertheless, the understanding of the effect created by lockdown on behavioural and emotional-mood domains in children and adolescents with ADHD poses great challenges based on the following considerations. First, negative mood and behavioural impairment are often expressed by most patients with ADHD with different degrees of severity, independently by the distress caused by Covid-19; as a consequence, we might expect that the impact of confinement on ADHD patients, will be defined by the changes in severity degree with respect to the previous status. Second, considering that the general functioning of children with ADHD is highly dependent on environmental context (Purper-Ouakil et al., 2004), we could expect that children and adolescents emotional-mood and behavioral variations during Covid-19 could represent a critical indicator of the change with respect to previous lifestyle. Therefore, the aim of the study is to examine the effect of lockdown on children and adolescents with ADHD based on the changes in severity degree of their emotional-mood state and behaviours with respect to previous condition, and to individuate which ADHD patients will result mostly vulnerable to the restriction condition. Subjects An anonymous online survey to be completed by parents was appointed for the study in order to evaluate the effect of the lockdown on emotional-mood states and behaviours of children and adolescents with ADHD. All parents were informed of the survey through the National ADHD Family Association website, for a limited time window (from June 4 to June 21, 2020), targeting patients with ADHD from 5 to 18 years old. The families were in total and partial lockdown for 88 days at the time of the survey. Before accessing the survey, parents were asked to read the written consent form and to agree to participate in the study. Informed consent represented an obligate field for advancing to the compilation of the questionnaire. All ADHD patients were followed and diagnosed by a child and adolescent psychiatrist of the Child and Adolescent Mental Health Services before the survey and the parents provided this information since they were registered in the National ADHD Family Association with a certified ADHD diagnosis. The Italian child and adolescent psychiatrists follow the ADHD Italian guidelines (that reflect international guidelines of the American Academy of Child and Adolescent Psychiatry) for clinical diagnosis and the obligatory protocol of the Italian Superior Health Institute. The total sample is to be considered as representative for the entire Italian territory with the participation of all regions, 20 metropolitan cities and 78.3% (72/92) Italian provinces. Data reported in this study were part of a wider research project designed with multiple purposes regarding the psychological impact of home confinement in Italy. There was no monetary or credit compensation for participating in the study. The study was approved by the Ethics Committee of the Department of Developmental and Social Psychology Sapienza University and was conducted in accordance with the Declaration of Helsinki (October 2008). Procedures The questionnaire relative to this study included 11 items indicative of most common negative emotional/mood (sadness, boredom, little enjoyment/interest, irritability, temper tantrums, anxiety) and disruptive behavioral problems (verbal and physical aggression, argument, disobedience/opposition, restlessness). We selected 11 questions and format of the items among emotional and behavioural dimensions of validated Italian version of Child Behavior Checklist (CBCL) 6-18 questionnaire (Achenbach and Rescorla, 2000;Frigerio, 2001). We choose these 11 items as the most appropriate to evaluate the psychological and behavioral effect of confinement condition in children and adolescents. Parents were requested to choose a single response defining severity degree based on the frequency per week by which their children/adolescents expressed each behavioural and emotional-mood dimension (1-2 time/week (low degree); 3-4 time/week (moderate degree); 5-7 time/week (severe degree), absent) before and during the confinement for

Covid-19 (Table 1). For the transitory period of the confinement (two months in Italy), we have considered that the expression of these problems in terms of frequency/week could be easier for parents in order to evaluate the severity parameter. Statistical analyses The responses of parents in the single question were modified in categorical variables (1= yes-0 =no) for each of the selected evaluation of severity degree. In order to identify which dimensions increased or decreased in severity under restriction, or which dimensions, not present before the lockdown, were expressed with one of three defined severity degree only during the lockdown (ODL), we have recoded the data of each participant on the base of severity changes with respect to previous condition. McNemar nonparametric chi square test for dependent samples was performed to compare: 1) each emotional-behavioural dimension with the same severity degree "before-during" lockdown; 2) worsening vs. improvement during the lockdown. Chi square was performed to compare children vs. adolescents on each mood and behavioural dimension. Statistical significance is set at a nominal two-tail P<.05, unless otherwise specified. Statistical analyses were conducted using SPSS software release 17.0 (SPSS INC, Chicago, Illinois). The questions were repeated in two separated items. Family economic status of the whole sample was high in 3%, middle in 75.2% and low in 21.8%. Mothers were the main compiler of the survey (88.6%). As for the education level the majority of participants have a graduate (29%) or high school degree (52.6%); middle school was represented in 17.9% and elementary school in 0.5%. Moreover, the family composition was reported as follows: parents with one offspring member: 31.3%; parents with two offspring members: 51.9%; parents with three offspring members: 13.3%; parents with four or more offspring members: 3.5%. Parents reported that 441 ADHD patients were taking medication, while 551 did not take medication before and during the lockdown. Changes in frequency of each emotional mood and behavioural domain before and during the lockdown During the lockdown we observed a significant decreased frequency of mood and behavioural problems expressed with low severity degree in both children and adolescents, with the exception of little enjoyment/ interest in children and physical aggression in adolescents (Table 2). In parallel, under restriction, we found an increase in frequency of the boredom, in temper tantrums and little enjoyment/interest domains expressed with moderate severity degree in both ADHD age groups; moreover, children also showed an increase in percentage of sadness and adolescents in physical aggression. No significant differences were found in both age groups in restlessness, opposition, verbal aggression, argument, irritability and anxiety dimensions. During the lockdown, children and adolescents with ADHD with high severity degree showed a different trend: we found in children an increase in percentage in almost all dimensions with the exception of restlessness and opposition while in adolescents we observed an increased percentage only in boredom, temper tantrums, little enjoyment/interest and argument. Changes in severity degree of emotional -mood and behaviours dimensions during the lockdown Children and adolescents with ADHD with previously low severity degree showed the lowest percentage of stability (no change) beforeduring the lockdown with significant fluctuations toward moderate and severe degree in all the dimensions, mainly in boredom, temper tantrums, little enjoyment/interest, argument, restlessness, irritability and sadness (Fig. 1). During the lockdown, these fluctuations contribute to explain either the decrease or the increase of frequencies in examined domains in ADHD patients with low and with high severity degrees, observed in the comparative analyses in Table 2. Moreover, we found a percentage > 20% of children and/or adolescents with ADHD that started to express little enjoyment/interest and physical aggression, and sadness and boredom with low severity degree during the lockdown (Fig. 1). Children and adolescents with ADHD with moderate degree showed more stability (no change) before-during the lockdown; nevertheless, we also registered a change toward the lower severity degree in all the dimensions with the exception of boredom among children. Moreover, we continued to find a percentage around 20% of children and/or adolescents that, during the lockdown, started to express boredom, sadness, argument and overall little enjoyment/interest with moderate severity degree (Fig. 2). Children and adolescents with high severity degree reported the major stability before-during lockdown with rates between 52% and 72% in boredom, temper tantrum, restlessness oppositional and verbal aggression and, in physical and verbal aggression behaviours among children and adolescents respectively. In parallel, in children and adolescents that have a severe degree, only during the lockdown (ODL), we observed a trend toward a lower severity in all dimensions but boredom, little enjoyment/interest and, only among children, sadness (Fig. 3). When we compared worsening and improvement relative to each domain during the lockdown, children and adolescents with previous low severity degree, showed significant worsening in almost all dimension with exception of anxiety and, only among adolescents, of sadness. In particular, boredom and little enjoyment/interest and temper tantrum and irritability showed a difference in percentage between worsening-improvement > 30% until 50% in both age groups. Moreover, a difference in percentage >30% was found in sadness and restlessness and verbal aggression among children and in argument among adolescents (Table 3). Under restriction, children and adolescents with previous moderate severity expression continued to report high rates of worsening in boredom and little enjoyment/interest, although with significant values only among children. Moreover, adolescents showed significant higher rates of improvement in restlessness and irritability. Children and adolescents with previous severe degree, during the lockdown, showed a significant improvement in opposition, restlessness irritability and argument; however, they continued to report higher rates of worsening in little enjoyment/ interest and equivalent rates of worsening-improvements in boredom. Inter-ADHD age groups comparison showed a significant difference in sadness among children with low severity degree while those with moderate degree reported high percentage in boredom and temper tantrum and irritability than adolescents. Discussion The current study in ADHD patients, aimed to examine the impact of the lockdown on emotional-mood and behavioural domains based on changes in severity degree with respect to previous condition and to individuate ADHD patients who resulted more vulnerable to lockdown experience. During the lockdown, a first evidence of the study consists in finding different profiles of the emotional-mood states and behavioural dimensions among ADHD patients with distinct severity degree. These differences were found either in the percentage of stability of degree before-during the lockdown than in the changes. During the lockdown, ADHD with low severity degree in the mood-behavioral domains showed the least before-during stability, the largest scale of fluctuation and significant worsening in almost examined domains. The understanding of this general worsening would require further investigations, although emotional lability and mood instability, and adaptive disfunctioning are closely associated and common features reported among ADHD patients (Anastopoulos et al., 2011;Barkley andFischer, 2010;Brotman et al., 2006;Sobansky et al., 2010). Because similar results were reported in Lee et al. (2020) study on children and adolescents during the lockdown, it is presumable that the sudden interruption of friendly relationships or of the opportunities of pleasant activities, in patients with lower dis-functioning, could have caused a critical adaptation problem with consequent outbreaks of emotional-mood status and of the behaviours. In general, these results support the consideration that excessive and rapid changes, in affective-motivation-arousal, and/or behaviors represent a critical expression of distress to environmental mismatch independently by age and by mental health state. Conversely, ADHD patients with high and moderate severity degree, although maintained a higher stability of their severity degree before and during the lockdown, showed important rates of improvement in several emotional mood and behavioural dimensions. Due to fact that similar results were found in both age ADHD groups, they suggest that the restriction could have represented, for some children, a protective Table 2 Comparison of emotional-behavioral dimensions expressed with the same severity degree before and during lockdown in children and adolescents with ADHD. condition from common social stressors, such as fewer friendships, bullying, victimization, and rejection of peers (Hoza, 2007;Nijmeijer et al., 2008) or school time constraints (Bobo et al., 2020;Chawal et al., 2020). The findings, in confirming the strong relationship between severity degree and stability of behavioural impairment, highlighted that these ADHD patients are improvable when environmental context is more flexible and responsive. Our results are consistent with those of Zhang et al. (2020), that reported significant relationship between improvements in the online study at home with longer time of study and a decrease of ADHD symptoms. To a closer exam relative to each domain, the findings highlighted that, the lockdown has overall caused, in both age groups, a worsening of boredom, in particular among ADHD patients with low and moderate severity, and of the enjoyment/interest, independently by severity degree, although with significant results exclusively among children. The first finding was expected because boredom is closely linked to ADHD condition, susceptible to important fluctuations as well as a critical hallmark of distress caused by environmental condition. The results are consistent with previous studies (Zhang et al., 2020), that reported a Fig. 1. Changes in Children and Adolescents with Low Severity Degree % of Children (C) and Adolescents (A) with previous low degree that maintained (no change), increased or lowered their severity degree or started to express emotional-mood behavioural problems with low severity degree (ODL) during lockdown. Fig. 2. Changes in children and adolescents with moderate severity degree. % of Children (C) and Adolescents (A) with previous moderate degree that maintained (no change), increased or lowered their severity degree or started to express emotional-behavioural problems with moderate severity degree (ODL) during lockdown. positive association between negative mood states and worsening of ADHD symptoms during the lockdown. Nevertheless, the co-occurrence of boredom and little enjoyment/interest in activities oriented versus a closer relation to lockdown condition. It is reported that the lockdown has caused a significant increased difficulty in keeping track of time (Cellini et al., 2020) providing prolonged boredom times . The psychological perception of the time is usually marked by the duration of the activities, by the variability of information and stimulations, by dynamic environments and contextual changes. In agreement with our findings, it is well documented that the reduction of pace of time promotes an increase of boredom and of the under-motivation state, sometime associated to a general decrease of well-being reflecting on adaptive behaviors. (DanZakai, 2014). The high percentage of ADHD in both age groups that, only during the lockdown, showed boredom and little enjoyment/interest, confirms these domains as the most affected by the restriction. In particular, among children, little enjoyment/interest was the only mood dimension that significantly worsened independently by severity degree. Conversely, despite the risk of Covid-19 infection and according with some studies (Bobo et al., 2020), ADHD patients didn't result particularly affected by anxiety in both age group and, independently by severity degree, showed equivalent rates between worsening and improving. With the exception of ADHD patients with low severity degree, we registered the highest rates in stability degree before-during the lockdown in physical and verbal aggression, oppositional behaviours, and restlessness and an important percentage of patients of both age groups that started to express argument with significant severity parameters, during the lockdown. According with other studies (Bobo et al., 2020), our findings confirmed that the condition of restriction could promote conflict within family environment. To a general examination, the trend of worsening and improving was substantially similar in both age groups, although children resulted more susceptible to mood fluctuations than adolescents did. It is well documented that ADHD patients showed less flexibility in the use of coping strategies for dealing with stressful situations than control subjects as shown by Babb et al. (2010). The same authors reported a higher coping flexibility in older (10-11 years) vs. younger (7-8 years) typically developing children while this difference was not found in ADHD children with the same age. Since studies show a developmental delay of brain areas implied in the executive functioning among children with ADHD (Shaw et al., 2007) we can assume that the difference in coping flexibility might become evident at later ages. Therefore, the mood patterns differences found between children and adolescents with ADHD in our sample could be linked to the higher flexibility of adolescents in the use of emotional coping strategies. Considering the representativeness of the sample, these findings represent an important goal of the study, in defining ADHD emotionalmood and behavioural responses to distress caused lockdown. Conclusion The

current approach provided a punctual individuation of domains affected and the indirect benefits produced by restriction condition in children and adolescents with ADHD. In view of reinstatement of regular routine, the major challenge after the pandemic will be to deal with its sequelae, also considering the risk of going back to previous severity condition in patients that improved during the lockdown. Moreover, it is our opinion that these findings could provide implication for the clinical practice and interventions independently from the specific aim of the study. The current study has some limitations that must be acknowledged. Although the survey was conducted after few days the end of lockdown and in condition of ongoing yet restriction, we cannot exclude a memory bias of the parents. We have not evaluated the relationship between impairment of the examined domains and ADHD severity, focusing on the problems reported as most concerning for parents. Moreover, we did not examine the relationship with the medication status, since parents did not answer to a specific question about the access to medication during the lockdown. Another limitation is that we did not include the ADHD diagnosis of parents as covariate/confounder in our analyses. Finally, considering the richness of the survey, we decided to select 11 items of the CBCL, as the most appropriate to evaluate the psychological and behavioral effect of confinement condition, in order to ensure the best return rate. Future extensions of the present study will allow to uncover the relationships with acute stress symptoms, with the comorbidity profiles of the ADHD patients, as well strategies adopted by parents to manage the Fig. 3. Changes in children and adolescents with high severity degree. % of Children (C) and Adolescents (A) with previous high severity degree that maintained (no change), lowered their severity degree or started to express affective-behavioural problems with severe degree (ODL) during lockdown. child-adolescent's problems. Authorship responsibility Each author made a substantive intellectual contribution to the study. • Maria Grazia Melegari: conceptualization and study design; data collection and interpretation; preparation and revision of the manuscript; approved the final manuscript as submitted • Martina Giallonardo: data collection and interpretation; revision of the manuscript; approved the final manuscript as submitted. • Roberto Sacco: data analysis and revision of the manuscript; approved the final manuscript as submitted. • Lavinia Marcucci: data collection, preparation and revision of the manuscript; approved the final manuscript as submitted • Silvia Orecchio: data collection; preparation and revision of the manuscript; approved the final manuscript as submitted. • Oliviero Bruni: conceptualization and study design, data collection, interpretation; revision of the manuscript; approved the final manuscript as submitted. Funding The authors received no financial support for the research, authorship, and/or publication for this article. Declaration of Competing Interest The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Socioeconomic status of practice location and Australian GP registrars’ training: a cross-sectional analysis Background Socioeconomic status (SES) is a major determinant of health. In Australia, areas of socioeconomic disadvantage are characterised by complex health needs and inequity in primary health care provision. General Practice (GP) registrars play an important role in addressing workforce needs, including equitable health care provision in areas of greater socioeconomic disadvantage. We aimed to characterize GP registrars’ practice location by level of socioeconomic disadvantage, and establish associations (of registrar, practice, patient characteristics, and registrars’ clinical behaviours) with GP registrars training being undertaken in areas of greater socioeconomic disadvantage. Methods A cross-sectional analysis from the Registrars’ Clinical Encounters in Training (ReCEnT) study. ReCEnT is an ongoing, multi-centre, cohort study that documents 60 consecutive consultations by each GP registrar once in each of their three six-monthly training terms. The outcome factor was the practice location’s level of socioeconomic disadvantage, defined using the Index of Relative Socio-economic Disadvantage (SEIFA-IRSD). The odds of being in the lowest quintile was compared to the other four quintiles. Independent variables related to the registrar, patient, practice, and consultation. Results A total of 1,736 registrars contributed 241,945 consultations. Significant associations of training being in areas of most disadvantage included: the registrar being full-time, being in training term 1, being in the rural training pathway; patients being Aboriginal or Torres Strait Islander, or from a non-English-speaking background; and measures of continuity of care. Conclusions Training in areas of greater social disadvantage, as well as addressing community need, may provide GP registrars with richer learning opportunities. Background Socioeconomic status (SES) is a major determinant of health, and of central importance to the work of health care providers, including general practitioners (GPs, family physicians) [1,2]. At the individual and area of residence level, greater socioeconomic disadvantage is associated with a disproportionate burden of disease, with higher rates of illness across all categories, particularly chronic diseases and multi-morbidity, and within a wider context of the concomitants of lower SES, including under-or unemployment, insecure housing, and poor social supports [3][4][5]. There are strong associations of residing in an area of greater socioeconomic disadvantage and disease risk factors, including smoking and poor nutrition, as well as lower uptake of preventative care including immunisations and health-screening [4][5][6][7]. An adequate supply of primary care physicians attenuates disparities in health across socioeconomic status [2]. But people in areas of greater socioeconomic disadvantage have increased difficulty accessing primary healthcare, including longer wait times and shorter consultation times, resulting in lower rates of patient enablement and patient satisfaction [3,[8][9][10][11][12]. This represents a manifestation of the 'inverse care law' , where workforce shortages and maldistribution see those with the highest need of healthcare receiving the least care [9,13]. Thus, addressing SES-related health inequities relies on an ongoing workforce of adequately educated and trained general practitioners (GPs). The training period for GP registrars (specialist vocational trainees in general practice) provides an opportunity to influence their future work practices as GPs, and can assist in preparing GPs to respond to the medical, psychological and social needs of the most socioeconomically disadvantaged within Australia's evolving primary health care system. Clinical experience is fundamental in the adequate preparation of GP registrars for the complexities and challenges of future independent practice. Structural changes in junior hospital doctor clinical experience can limit pre-vocational exposure to factors critical to care of disadvantaged populations, including the comprehensive management of patients with chronic disease and multimorbidity. This makes it more important that registrars gain adequate exposure during vocational training [14]. It is therefore plausible that GP registrars training in areas of lower SES may benefit from a richer training experience, with increased exposure to higher levels of multimorbidity and more complex medical and psychosocial patient presentations. GPs practising in areas of socioeconomic disadvantage encounter higher rates of complex multi-morbidity and chronic disease, and may have a greater engagement in promoting preventative health care (through screening for biological and behavioural influences on health) [3]. However, the potential educational benefits from training in areas of socioeconomic disadvantage have not been well-established or explored. In addition to preparing registrars clinically, it is argued that government-funded GP training organisations bear a social obligation to acknowledge, and redress the inequality in healthcare across socioeconomic areas [15,16]. Promoting training in high-needs areas provides both short-and long-term benefits [16]. By training in areas of socioeconomic disadvantage, GP registrars contribute to the current GP workforce in often-underserved areas, while obtaining a real world orientation to their social responsibilities as GPs. The current exploratory study aimed to a) characterize GP registrars' practice location by level of socioeconomic disadvantage, and b) to establish associations of training in areas of greater socioeconomic disadvantage; including registrar, practice, and patient characteristics, and registrars' clinical behaviours. Methods This cross-sectional analysis took place within the Registrars' Clinical Encounters in Training (ReCEnT) study. ReCEnT ReCEnT is a cohort study of individual registrars' in-consultation clinical and educational experience. The complete methodology is described elsewhere [17]. Briefly, GP registrars collect data once at approximately the midpoint in each of their three six-month mandatory general practice training terms, capturing demographic data, diagnoses, investigations/management, and educational training aspects of 60 consecutive patient consultations. The project is an intrinsic element of registrars' training, and is compulsory [18,19]. Registrars may also provide informed written consent to their data being used for research purposes. From 2010 to 2015 it was conducted in Regional Training Providers (RTPs) across five of Australia's six states and, from 2016 (after a reorganization of Australian GP vocational training), in three Regional Training Organizations (RTOs) in three Australian states and the Australian Capital Territory. The number of registrars from participating RTPs/ RTOs consenting to use of ReCEnT for research purposes determined the sample size for this study. Patient and Public Involvement Patients or members of the public were not involved as participants in this study. Outcome factor The outcome factor was a measure of the registrar's practice location level of socioeconomic disadvantage. Practice location postcode was used to define the practice Socio-Economic Index for Area Relative Index of Disadvantage (SEIFA-IRSD) which we determined to be the most appropriate of the SEIFA indexes for this research question [20]. The SEIFA -IRSD summarises a range of social and economic variables of an area to provide an index of relative disadvantage. While low income is the strongest indicator of disadvantage, additional variables include employment type/unemployment, education, rent repayments, disability, internet connection, and household relationships such as single parenting, separation, and divorce [20]. All GP training practices who to have participated in ReCEnT were ranked by SEIFA-IRSD. The ranked-by-SEIFA-IRSD practices were categorized to form five quintiles, and then stratified so the SEIFA-IRSD quintile of greatest disadvantage was compared to the other four quintiles. Independent variables Independent variables related to the registrar, patient, practice, consultation, and consultation outcomes. Registrar variables included age, gender, full-time/parttime status, training term, place of medical qualification (Australia or International), training pathway, non-English speaking background, and whether the registrar had worked at the practice before. Practice variables included practice size (number of full-time equivalent GPs, with practices with less than five GPs categorised as small), geographic location (rurality) (using practice postcode to define Australian Standard Geographical Classification-Remoteness Area, ASGC-RA) [21], training region, and bulk-billing policy (whether consultations are free to the patient). Patient characteristics included age, gender, and whether the patient identified as Aboriginal and/or Torres Strait Islander, was from a non-English speaking background, and was a continuing patient or was new to the practice, or to the registrar. Consultation characteristics included consultation duration, number of problems/diagnoses managed, and whether the registrar sought information or assistance during the consultation (from their supervisor/ trainer, from a specialist, or from hard-copy or electronic sources), whether the problem was classified as a chronic disease [22], if any procedures were performed, and if the patient was seen by a practice nurse. Consultation outcomes included whether any imaging or pathology tests were ordered, whether any follow-ups were arranged, if any medications were prescribed, if any referrals were made, and if the registrar generated any learning goals during the consultation. Statistical Analysis This was a cross-sectional analysis. Analysis was performed on 16 rounds of data collected between 2010 and 2017. Individual regions contributed 2 to 17 rounds of data depending on when they entered the project and on continuity/discontinuity across the 2015-2016 restructure of Australian GP vocational training. The unit of analysis was the consultation. The proportion of consultations in the lowest SEIFA-IRSD quintile was calculated with 95% Confidence Interval (CIs). Univariate logistic regressions were undertaken to examine the relationships between the outcome factor and independent variables. Variables with a P-value of <0.20 were considered for inclusion in the multivariable logistic regression model. Logistic regression was used within the generalised estimating equations framework, to account for repeated measures within registrars. Once multivariable models were fitted, model reduction was assessed. Covariates not reaching p<0.20 in the multivariable model were tested for removal from the model. If the covariate's removal did not substantively change the resulting model (defined as any covariate in the model having a change in the effect size (odds ratio) of greater than 10%), the covariate was removed from the final model. To examine different facets of our research question, three models were built, each with 'quintile of greatest socioeconomic disadvantage' as the dependent variable. To examine the associations of a consultation being conducted in the greatest disadvantage quintile (i.e., lowest SEIFA-IRSD quintile), patient, practice and registrar independent variables were included in a multivariable regression

model. To examine how consultations conducted in the area of greatest disadvantage quintile differ from other consultations, the above variables were included in a second multivariable model along with the following additional variables: consultation duration, the number of problems addressed during the consultation, if chronic conditions were managed, and if any sources of information or advice were consulted. To examine how outcomes of consultations in areas of the quintile of greatest disadvantage compared to those of other consultations, all variables from the previous models were included in a final multivariable model along with the following additional variables: if procedures were performed, follow-up organised, and whether learning goals were generated. The rationale for building three models was that associations of a registrar's consultation being conducted in the lowest SEIFA-IRSD quintile practice will include patient, registrar and practice factors, but evaluation of these associations may be compromised by inclusion in the multivariable model of factors operating once the consultation is progressing. Similarly, evaluation of the content of the consultation may be compromised by the inclusion in this model of outcomes arising from the consultation. Of the 29 covariates of interest, 24 were considered for inclusion in the multivariable model. However, 'region' and 'rurality' , were subsequently removed from the model, due to high correlation of these variables with each other and with the outcome, causing instability of parameter estimates due to data sparsity and collinearity. Variables were considered statistically significant if the P-value was <0.05. Ethics approval Ethics approval was from the University of Newcastle Human Research Ethic Committee, Reference H-2009-0323. Characteristics of the registrars and their practices are shown in Table 1. Characteristics associated with training in a practice in the lowest SEIFA-ISRD quintile versus the four highest quintiles are presented in Table 2. Results of univariate and multivariable logistic regression models are presented in Table 3. Multivariable associations Statistically significant (at p<0.05 level) registrar-level multivariable associations of conducting a consultation in the lowest SEIFA-IRSD quintile practices included: the registrar working full-time ( There were no significant multivariable consultationlevel or consultation-outcome associations. Main Findings and comparison with existing literature While existing research into established GPs and GP registrars often examines socioeconomic status as a study variable, few studies focus on SES as the outcome factor. This study is therefore unique in highlighting key considerations for GP registrars training in practices in areas of socioeconomic disadvantage. In this study, registrars who undertook training in areas of greater socioeconomic disadvantage had greater exposure to patient populations known to have a greater burden of chronic disease, multi-morbidity, and complex social needs. This included patients of Aboriginal and/or Torres Strait Islander background, and patients from a non-English-speaking background [5]. This is consistent with patient demographics of areas of greater socioeconomic disadvantage in Australia [23]. GP registrar's continuity of care was also a theme of the results, with patients in more disadvantaged areas being more likely to have seen the registrar previously. The patient was also less likely to be new to the practice. And registrars in disadvantaged area practices were more likely to have worked at the practice previously. This cluster of findings suggests continuity of care. Maintaining a relationship with a GP is an essential element to patient engagement and satisfaction [12], and is also recognised as of significant benefit for the registrar in preparing them for independent practice [24]. Registrars were also more likely to be on the rural pathway, which we have found previously to provide a more diverse clinical experience [25], and there was some evidence (p=0.057) for an association with more problems seen per encounter than in areas of higher SES. These findings, along with the opportunity to treat patients from higher-needs groups and increased continuity of care, support the potential for a clinically richer training experience for registrars at practices in of greater socioeconomic disadvantage. We also found that GP registrars in their first training term were significantly more likely to work in areas of greater disadvantage than those in later terms. This suggests an educational 'immersion' in the rich training experience that these practices may offer. This may be similar to registrars being exposed to the rich learning environment of rural practice at the beginning of their training [25]. Strengths and limitations A strength of this study is the generalisability of results. The ReCEnT study covers all categories of rurality and includes a comprehensive mix of SES areas. The large sample size and a high response rate [26], as well as a A possible limitation of this study is the use of practice, as opposed to patients' , geographic level of disadvantage. However, while practices themselves may see a mix of patients from all SES backgrounds, the focus of this study was to highlight the experiences of registrars within these practices overall, rather than with any individual patient. A further limitation is that we have data only on the content of individual consultations. We do not have data on past medical or social history, or medicine regimens. Implications for policy and practice There were associations identified in this study that indicate a richer training experience for GP registrars who train in practices located in areas of greater socioeconomic disadvantage. These include associations with working with patients from groups with clinical complexity, and with markers of greater continuity of care. This may contribute to registrars' learning, as well as helping meet the current health care needs of disadvantaged areas (noting that registrars comprise 13% of Australia's general practice workforce (by headcount) [27,28]). An implication is that registrars should be strongly encouraged to train in lower SES-areas. It might be thought that early exposure to the richness and complexity of medicine in disadvantaged areas may, in itself, encourage registrars to continue to work in these areas, including post-Fellowship. Our findings, however, of an association of later training term with less disadvantaged practice setting may suggest that registrars may move away from more disadvantaged practices during training (though our cross-sectional study cannot establish temporal patterns in registrars' practice location). If this is so, it may be a concern that this trend could continue into registrars' post-Fellowship choices of practice location. This may suggest attempts to address areas of high need, such as low socioeconomic disadvantage via GP vocational training, may be limited in rebalancing health equity and workforce issues beyond the immediate effect of vocational training time. An additional consideration is that while we have found evidence of training in lower-SES areas providing a rich educational environment, this may also represent a clinically challenging environment (especially the challenges of complex multimorbid disease in socially complex contexts). That we have found that registrar experience in lower-SES areas is 'front-loaded' earlier in training may have implications for vocational GP training. Within the apprenticeship-like model of Australian GP vocational training, supervisor in-practice oversight of registrars' learning and practice is concentrated to greater support early in training. Given that structural approaches to redress the ratio of first-term to later-term registrars in disadvantaged practices may be difficult to implement, our findings suggest that front-loading of supervisory support could be even further resourced in disadvantaged areas. Implications for future research Future research is required to understand in greater detail the experiences of registrars in areas of greater socioeconomic disadvantage, and what influence this has on their future placement and practice location choices. Careful consideration needs to be given to less experienced registrars training in areas of socioeconomic disadvantage, and what supports may be required if this pattern continues. Further research of this area would assist in understanding the experiences of those working in disadvantaged areas, as well as the impact practice location SES has on the learning outcomes for GP registrars. Conclusion Our findings suggest that GP registrars training in areas of greater socioeconomic disadvantage are exposed to a broader range of clinical and educational experiences and learning opportunities. Registrars should consider undertaking training in these areas to take advantage of the range of these experiences. The continued support of registrars working in these areas, and the encouragement of more senior registrars to work there, also has the potential to assist in addressing health inequity experienced within these communities. The Distribution of Facial Profile Photogrammetry of High School Students in Medan One of the main objectives of orthodontic treatment is to improve dental and facial esthetics. The understanding of soft tissues of the face in relation to the dentoskeletal tissues is essential in esthetic orthodontic treatment. Study of soft tissue facial profile by photogrammetry provides better information on the facial morphology. This is a descriptive study with a cross sectional approach using level stratified-cluster sampling method. The subjects of this study are 200 high school students in Medan. Lateral photographs of each subject were taken and their facial profile was measured. Results: The facial profile of high school students in Medan are convex 92.5%, straight 5.0%, and concave 2.5%. There were no significant differences in the facial profile of high school students in Medan based on gender. The facial profile of high school students in Medan is generally convex. Keywords–high school students, facial profile, INTRODUCTION Determination of the lateral facial profile is one of the many orthodontic examinations conducted for the purpose of diagnosis and treatment planning. The value of facial profile represents the facial form and these values differ in each individual [1,2]. Despite its inability to provide complete orthodontic information, determining the facial form enables evaluation of jaw malrelation [3]. Orthodontic treatment aims to arrange the dentition to achieve facial balance as well as restoring the relationship of occlusion [4]. Achieving balance between the teeth and the facial profile is also a goal of orthodontic treatment [5]. Thus, the correction of the facial profile is an important criteria of successful treatment. Sarver et al [5] stated that for several decades, doctors had tried to determine an ideal facial proportion. The study continues to this present-day and remains an interesting topic for orthodontists as a standardized guideline for aesthetic assessment has yet been established. Facial soft tissue analysis can be done by several methods; direct method of soft tissue measurement, lateral cephalometric radiographs, and photogrammetry [6,7]. Photographs provide a good assessment of harmony between the external craniofacial structures, including the soft tissues, in addition to providing reliable measurements. Through photogrammetric analysis of the facial profile, proportionality, angular and linear measurements can be obtained [8]. Soft tissue profile standards using photogrammetry have been reported for North American, Spanish, Indians, Brazilian Caucasians, Croatians and Turkish population [9]. According to Graber there are three forms of lateral facial profile that can be viewed laterally, namely; concave, straight, and convex [1,10]. Linden [11] suggests that the facial profile of males generally presented a straighter profile compared to those of an adult female. Studies of facial profile that had previously been carried out include research by Al-Barakati [12] which states that the population of Saudi adults have more convex facial profiles than the population of Europe and America as characterized by the angle of facial convexity while among the Saudi population itself, the facial profile of women was found to be relatively more straight compared to men. Zylinski [13] on children and white young adults showed that their facial profile is relatively straight rather than children due to the upper and lower lips in young adults being more retrusive. II. MATERIALS AND METHODS This study was conducted in a descriptive manner with a cross sectional approach to obtain the facial profile of high school students in Medan. The population of this study were high school students in Medan with a total of 116.038 students from 18 government high schools and 138 private high schools in Medan based on the data obtained from the Sumatera Utara Education Office for the year of 2006. The sampling method used was level stratified-cluster with a total sample of 200 high school students. The inclusion criteria for this study were subjects with an age range of 15-18 years old, neither previously nor currently undergoing any orthodontic treatment, Angle Class I molar relationship with normal overjet and overbite (2-4 mm), mild crowded anterior teeth and diastema ≤ 2 mm, not wearing dentures, has a complete set of permanent dentition (except third molars), competent lips, and possessing no facial asymmetry. The exclusion criteria were high school students with a history of fracture/trauma, history of plastic or

maxillofacial surgery or refusal to participate in the study. A. Procedure of study 1. Ethical Clearance and Informed consent: Ethical clearance was approved by submitting permission to the Health Research Ethical Committee of Sumatera Utara c/o Medical School, University of Sumatera Utara. After approval, data of the study was gathered. Informed consent, in the form of written signature, was obtained by providing explanation regarding the aim and benefits of the study before performing examination and photogrammetry. 2. Sample screening: Subjects of the study were high school students between the ages of 15 to 18 years old in Medan. 3. History taking and individual data: Subjects were then examined intraorally (with a dental mirror, dental explorer, and tweezers) and questioned according to the inclusion and exclusion criteria. 4. Lateral photographs of the subjects were taken and printed. B. Printing and measuring photo results After photos were printed in 3R size, measurements according to the Graber's facial profile method (Gl-Pog reference line as a guide for determining facial profiles) was performed with the following steps ( Figure 1 Before analyzing data of all sample profile photos, an intra-operator test was conducted by performing measurements on 10 profile photos, and repeating it the next day. If similar results were obtained, the validity of the measurements would be assured and the operator would be selected to analyze all sample photos. A. C. Processing and analyzing data Data of the study was processed by computerization. Data analysis begins with measurement of the respondent's photo results using reference points of facial profile photo according to Graber; the glabella, upper and lower lip contours, and pogonion with pencil. Measurements was performed on a transparent plastic sheet. The results obtained would determine the facial profile of respondent based on non-parametric method, and then be tested descriptively to analyze the facial profile. III. RESULTS This study was conducted on 200 high school students. The overall facial profile distribution of high school students in Medan based on photogrammetry were convex, with the percentage of 92.5% convex, 5% straight, and 2.5% concave ( Table I). The facial profile distribution of high school students in Medan based on gender showed that there were no significant differences between facial profiles of male and female high school students in Medan (Table II). IV. DISCUSSION According to Salzmann, orthodontic dentistry is a branch of science that studies the development and anomalies of the teeth and jaws that affect oral health, physical health as well as aesthetic and mental stability. Understanding the importance of oral, physical and mental health, as well as the aesthetic appearance of Advances in Health Science Research, volume 8 patient is necessary. The aim of orthodontic treatment according to Riedel is achievement of "Utility", "Beauty" and "Stability". Similarly, according to Jackson's Triad, orthodontic treatment goal is to maximize the function of the stomatognathic system, preserve balance of the teeth structure, skeletal and soft tissue as well as the aesthetic balance [14]. The success of orthodontic treatment is characterized by good facial appearance. The soft tissue is an important factor that determines the final appearance of the patient's face [12]. Soft tissue facial profile analysis plays an important role in orthodontic treatment planning. Orthodontic treatment according to the accepted hard tissue cephalometric criteria does not necessarily ensure that overlying soft tissues will be positioned in a harmonious manner and hence, may not result in an esthetic profile. Soft tissue of the face requires an independent appraisal in addition to the skeletal and dental analysis in order to deduce a comprehensive diagnosis and treatment planning of the face. Soft tissue cephalometric norms for esthetic profiles had been established by various researchers using cephalometric radiographs. Recently, many studies had been conducted on the photographic evaluation of soft tissue facial profiles in some racial groups [15]. Stoner [16] had started to analyze facial soft tissues using photographic records. Farkas [16] standardized the photographic technique by taking the records in natural head position (NHP). The reliability of the photogrammetry, especially for the measurement of the lips and mouth was assessed, strongly recommended and suggested that the benefits of photogrammetry can be increased by developing better techniques. The facial appearance is subjective as it is influenced by various factors such as race, ethnic, gender, sociocultural, and age [7]. Due to the differences in perception, clinicians should consider the suitable normal value of the ideal facial appearance based on the patient's race and ethnicity when planning a treatment [6,17]. Ideal facial appearance is influenced by the soft tissue that covers the skeletal tissue [16]. Facial profile is one of soft tissue component that has to be examined for orthodontic diagnosis [18]. Table 1 shows that the overall facial profile distribution of high school students in Medan was convex 92.5%, straight 5%, and concave 2.5%. The results of this study were similar to Kosoemahardja [6] that stated the soft tissue profile of Mongoloid race were more protrusive than of Caucasian. Heryumani stated that the facial profile tends to be convex on Javanese due to the nose and chin being less protrusive [6]. Table 2 shows no significant differences between male and female facial profiles among high school students in Medan. Heryumani [6] examined the facial sagittal proportions of males and females in Java and found that the nose's depth, lip to sagital nasal tip distance and chin to sagital nasal tip distance in Javanese males and females were convex. Mauchamp and Sassiouni [19] studied on effects of gender on the measurement of soft tissue facial profile as pogonion (Pog), subnasal (Sn), and the convexity face profile. Their study reveals that female Caucasoid facial profile was found to be more straight compared to males between the ages 20-25 years old,. According to Subtelny [20] due to the thickness of the soft tissue covering the bony tissue, males were found to be more convex than females. Hambelton [21] found that male facial profile less convex than female, claiming it was due to the growth of the chin in men forming an angle greater than females. Palestinian adults showed skeletal profile convexity and this value was increased significantly in men (5.6 ±1.8mm) compared with women (4.5 ± 2mm) [22]. Ioi et al [23] stated that Japanese facial profile was highly convex for both males and females as both genders have a tendency of retruded lip position. Reddy et al [24] after calculating the total angular facial profile (G-Prn-Pg), it was found that the population of northern India has a nose more aquiline causing the facial profile to be more straight compared to the Europeans, and by calculating the angle G-Sn-Pg, the facial profile of the female population of north India were more convex than men. Anibor and Okumagba [25] stated that the higher values shown by the Urhobo females in their Nasomental and Mentocervical values and the lesser values in the Nasofrontal and Nasofacial angles compared to males imply that the females have a more protruded nose, less prominent glabella and bigger chin on their faces compared to males. In a facial profile study by Ferdousi AM, et al., on Bangladesh Christian Garo population, the angle of facial convexity was found to be higher in females (169.26° ± 4.43°) than males (158.65° ± 12.17°), similar to that in the North Indian population and White European population. The higher convexity in females could be explained by the fact that in general the facial contours of females were softer than those of males, especially in the area of the nose, lips and chin [26]. Therefore, clinical investigations are recommended and greatly needed to obtain the facial profile according to all races and ethnics in Indonesia. A study to assess the impact of left atrial size reduction in outcome of the patients undergoing mitral valve surgery for mitral valve disease with left atrial enlargement in a tertiary care hospital Rheumatic heart disease (RHD) is a permanent sequela of rheumatic fever (RF). Globally, there are >15 million cases of rheumatic heart disease, with 233,000 deaths each year and 282,000 new cases per year. The overall prevalence of RHD in our country is estimated to be about 1.5-2/1000 in all age groups (total population about 1.27 billion) which suggests that there are about 2-2.5 million patients of RHD in our country. INTRODUCTION Rheumatic heart disease (RHD) is a permanent sequela of rheumatic fever (RF). 1,2 Globally, there are >15 million cases of rheumatic heart disease, with 233,000 deaths each year and 282,000 new cases per year. 3 The overall prevalence of RHD in our country is estimated to be about 1.5-2/1000 in all age groups (total population about 1.27 billion) which suggests that there are about 2-2.5 million patients of RHD in our country. 4 Rheumatic fever affects all the valves microscopically, but clinically significant disease is observed mostly in mitral valve. Rheumatic mitral valve disease (mitral stenosis or regurgitation) remains the common heart disease in developing countries. Mitral valve is involved in 99% of cases. Pathologically rheumatic MS leads to leaflet thickening, commissural fusion, chordal shortening and fusion. Rheumatic mitral regurgitation leads to LA enlargement along with rise in pulmonary arterial pressure. 5,6 The LA is far from being a simple passive transport chamber. It is highly dynamic and responds to stretch with the secretion of atrial natriuretic peptides. 7 The counterbalance of natriuretic, vasodilatation, and inhibition of the sympathetic and renin-angiotensinaldosterone systems allows partial restoration of fluid and hemodynamic balance. 8 Left atrial geometry and mechanical function exert a profound effect on LV function and cardiovascular performance. 9,10 Enlargement of the LA has been shown to be a reliable predictor of adverse cardiovascular outcomes. The LA diameter has also been shown to independently predict death in the general population. 9 In other populationbased studies, the association of LA enlargement with mortality has been attenuated when diastolic function, LV mass, or LV hypertrophy has been considered. 9,10 The prognostic implication of LA size has also been shown in high-risk subgroups, such as patients with acute myocardial infarction, atrial arrhythmia, LV dysfunction, or dilated cardiomyopathy, and patients undergoing valve replacement for aortic stenosis and mitral regurgitations. [9][10][11] The relationship between LA enlargement and stroke is complex. LA size has been shown to predict ischemic stroke in subjects without atrial fibrillation (AF) and mitral valve disease in the Framingham heart study. The LA size is an independent predictor factor for thromembolism. 12 AF is closely related with the LA enlargement. AF and LA enlargement cause several morbidities and mortalities. Stroke is the most feared complication of AF. 13,14 Most of the patients who undergo mitral valve surgery also have LA enlargement and chronic AF. 11,13,14 Mitral valve surgery alone mostly does not restore sinus rhythm or prevent recurrence of AF after surgery. The idea that mitral valve surgery alone will result in remodeling and atrial size reduction is considered wrong by most studies. [12][13][14][15][16] After atrial size reduction, sinus rhythm was restored in 77.3% of patients, whereas in the group without reduction it was restored only in 61.1% of patients. 17 Addition of LA size reduction to mitral valve surgery is effective in 63% of patients with chronic AF in reducing the risk of stroke and thromboembolic complications. 14,17 The concept of left atria size reduction was developed primarily to improve left atrial geometry and to reduce thromboembolic risk in patients with left atrial enlargement and mitral valve disease, left atrial size is critically important for the restoration of the sinus rhythm, and LA reduction alone may augment the maintenance of the sinus rhythm. There is increasing evidence that LA size is potentially modifiable with medical therapy, but whether LA size reduction in patients undergoing mitral valve surgery translates to improved outcomes remains to be established. This has been the basis of study to know the impact of LA size reduction in mitral valve diseases. Objective of the study were to study the impact of left atrial size reduction in patients undergoing mitral valve surgery for mitral valve disease with left atrial enlargement on clinical outcome and echocardiographic parameters. METHODS A prospective study was done at department of cardiovascular and thoracic surgery, Sri Venkateshwara institute of medical sciences, Tirupati from June 2012 and June 2013. During the study period a total of 40 study subjects were enrolled in our study hence these 40 study subjects were grouped

into two categories with 20 study subjects. The study population consisted of 40 study subjects with twenty patients in each group patients with rheumatic mitral valve disease with or without tricuspid valve disease with left atrial size enlargement who underwent mitral valve surgery alone and mitral valve surgery with left atrial reduction. Inclusion criteria Inclusion criteria included patients with rheumatic valvular heart disease with left atrial enlargement a) mitral stenosis b) mitral regurgitation c) mitral stenosis and mitral regurgitation d) MS/MR with tricuspid regurgitation. Exclusion criteria Exclusion criteria excluded patients with valvular heart disease undergoing mitral valve surgery along with aortic valve replacement, CABG and for other conditions. Patients were evaluated by history, physical examination, biochemical tests, ECG and transthoracic echocardiography. CAG was done for patients above 40 years of age according to guidelines. Diagnosis of valvular heart disease with left atrial enlargement was established by echocardiography and graded accordingly. Subjects with mitral valve disease and left atrial enlargement with or without tricuspid valve disease undergoing mitral valve surgery were included in the study. The present study included 40 patients. Among these subjects undergoing mitral valve surgery and left atrial size reduction were included in group A-LA reduction group. Subjects undergoing mitral valve surgery without left atrial size reduction were included in group B-No LA reduction group. Twenty consecutive subjects were taken in each group during the study period. Under GA CPB, mitral valve surgery and left atrial size reduction were done. The techniques of surgical reduction were lateral wall plication, dome plication, exclusion of left atrial appendage, resection of left atrial redundant wall and restoring. The choice of the technique was dependent on the assessment and at surgeon's preference. Surgical procedure All patients underwent median sternotomy to assess the heart. Pericardial trough was made. Heparin was given in the dose of 3 mg per kg body weight IV and the dose was monitored by activated clotting time (ACT). The aim was to achieve complete hemostatic paralysis. Pulmonary artery and left atrial pressures were recorded pre-and post-cardiopulmonary bypass. Aortic and bicaval cannulation was done and cardiopulmonary bypass was instituted. Surgeries were carried out under mild hypothermia (30-32°C). Aorta was cross clamped and warm ante grade intermittent blood cardioplegia was used for arresting the heart which was repeated at every 20 minutes. In addition, topical ice-slush was used for myocardial protection. The mitral valve was assessed in majority of the patients through left atriotomy. Right atrial trans septal approach was used for two patients who had concomitant tricuspid valve annuloplasty and in patients with small left atrium. Patients who had left atrial/left atrial appendage clot was removed and left atrial appendage exclusion was done. The method of replacement of mitral valve was decided preoperatively on the pliability and extent of calcification of the native mitral valve and the sub-valvular crowding. Modified mitral valve replacement with preservation of posterior mitral leaflet was carried out in all patients. Regular follow up was done at regular intervals within fifteen days after discharge, every month for the first three months, every two months for the next six months and every three months thereafter. Detailed history, physical examination and ECG, transthoracic echocardiography was done during follow up. Rhythm on ECG, LA dimension. LVEDD, LVESD, LVEF, RVSP, mitral valve gradients, tricuspid regurgitation jets were evaluated and compared between pre-op, immediate postop and follow up at six to twelve months. Any complications were treated and necessary drugs for anticoagulation were given with proper advice. The ethical approval was taken from the ethical committee of the institute. Statistical analysis The data was entered in MS excel and analyzed using SPSS V 21. The continuous variables were represented using Mean and Standard deviation and categorical data was represented in the form of frequencies and proportions and chi square test will be used to check for association between quantitative data. P value less than 0.05 is considered to be statistically significant. RESULTS The age range is 21 to 61 years with 60% of them between 30 to 40 years of age, 26.5% of them between 41 to 50 years, 12.5% of them more than 50 years and 1% of them less than 30 years of age. The mean age is 39.7±10.3 years. in the study group. The total no of males was 17 (42.5 %) and females were 23 (57.5%) in the study group. There were more females in the study groups. More patients belonged to low socio-economic status. Dyspnea was the predominant symptom in the study population. In group A, 2 were in NYHA class II, 14 patients were in NYHA class III and 4 patients were in NYHA class IV. In group B, 2 patients were in NYHA class II and 15 were in NYHA class III and 3 were in NYHA class IV. Post operatively, at six to twelve months there was improvement in group A, 10 were in NYHA class 1 and 10 were in class II. 10 patients improved from NYHA class III-IV to class II. In group B, there was improvement from class NYHA class III-IV to class II in 15 of the patients, 4 number of patients were in class I and 1 patient was in class III. The mean NYHA functional class improvement was significant in patients of both groups at sixth day and at sixth to twelve months follow up postoperatively. Total no. of patients were in AF pre-operatively in both groups. Two of the patients were in sinus rhythm in both groups. Following LA reduction, 2 patients reverted to sinus rhythm postoperatively while the rest continued to be in AF in group A. During follow up at six months one more patient reverted to sinus rhythm in this group. In group B, sinus rhythm was not restored. There was difference in conversion to sinus rhythm in both groups. During the early postoperative period, the persistence of AF in group B patients was more apparent than in group A. Patients were discharged from the hospital in 7 to 10 days. None of the patients had prolonged stay. In group A, a significant decrease in LA size was noted postoperatively which was not observed in group B. Between 6 and 12 months post-operatively, LA size reduction was prominent in group A with a decrease from 60.4±8.04 mm to 44.8±6.8 mm (p<0.01). The reduction of LA was more significant in patients with pre-operative LA dimensions over 60 mm. In group B, at six months of follow up, LA size had reduced from 56±6.12 mm to 51±5.1 mm, there was no further change in LA dimension with subsequent follow up, LA size remained same. Left atrial enlargement-electrocardiography Left atrial enlargement (LAE) is seen in patients with mitral and aortic valvular disease, IHD, hypertension and some cardiomyopathies. Typically, the p wave is bimodal in some leads and ± in V 1 with evident final negative mode. The diagnostic criteria of LAE are as follows: p wave with a duration 0.12 second especially seen in leads I or II, generally bimodal, but with normal height; diphasic P wave in V 1 with evident final negativity of at least 0.04 second of duration because the second part of the loop is directed backward due to LAE). These two criteria have good specificity (close to 90%; few false-positive cases) but discrete sensitivity (<60%; more false-negative cases). The ± p wave morphology in II, III, and VF with a p=0.12 second is very specific and has a high positive predictive value (100% in valvular heart disease and cardiomyopathies). However, it has a low sensitivity and low negative predictive value for LAE. Mitral valve gradients decreased in both groups after surgery. DISCUSSION All patients are rheumatic in etiology in both groups. The mean NYHA functional class improvement was significant in patients of both groups at sixth day and at sixth to twelve months follow up post operatively. The improvement was from NYHA class III and IV to class II or class I. There were no patients left in NYHA class III-IV at six to twelve months of follow up in group A. One patient remained in NYHA class III in group B. Özerdem et al in their study on Left atrial reduction by posterior wall plication combined with mitral valve surgery noted at follow-up 6 to 28 months after surgery 87% were in NYHA class I, 13% were in NYHA class II. Kutay et al observed that the patients with enlarged left atrium had similar functional class improvement. 18,19 Erdogan et al reported similar findings of improvement pre operatively from NYHA functional class III (65%), and 35% were in class II to NYHA class I at six months on follow up. 20 AF usually accompanies MV disease at the time of surgery, especially when the LA is enlarged. This is the main determining factor in the occurrence and maintenance of chronic AF. MV surgery does not result in relief of AF. During the early postoperative period, the persistence of AF in group B patients was more apparent than in group A. Following LA reduction 2 patients reverted to sinus rhythm postoperatively while the rest continued to be in AF in group A. During follow up at six months one more patient reverted to sinus rhythm in this group. Dzemali et al reported that left atrial size reduction affects cardiac rhythm in patients with chronic AF undergoing mitral valve surgery. 21 The addition of left atrial size reduction to mitral valve surgery was effective in restoring sinus rhythm in 19% at discharge and in 63% of patients with chronic AF, restoring predominant SR at one year postoperatively. Marui et al reported that by LA volume reduction using plication technique sinus rhythm restoration was significantly better in the volume reduction group at 12, 24 and 36 months of follow-up (p<0.05) Kutay et al, observed that during the early postoperative period after MVR, patients with enlarged LA persisted to be in AF while SR was restored in significant of patients without LA enlargement. 19,22 Johnson et al in their study on Plication of the enlarged left atrium at operation for severe mitral regurgitation noted post operatively conversion of AF to sinus rhythm. 23 20 Tonguç et al reported no differences in left atrial size between plicated and non-plicated patients. 24 Pande et al found a decrease in LA size after 5 years of follow-up following MVR in enlarged LA patients, but other studies have reported a reduction in LA size in the immediate post-operative period. 25 Right ventricular systolic pressure decreased significantly following surgery. It decreased from 79.5±24.7 to 47.3±11.9 mmHg during the 1 st six days after surgery. RVSP further decreased to 30.5±11.47 mmHg in group A at six months. In group B, there was reduction in RVSP from 74±26.7 to 44±9.9 mmHg during the 1 st 6 days after surgery, it further reduced to 41.05±12.04 mmHg at six months. On comparison, the decrease observed in the LA reduction group was more than the decrease observed in group B over 6 months. This difference observed was statistically significant (p<0.01). This is due to the reversible nature of pulmonary arterial hypertension in rheumatic mitral valvular heart disease. Limitations The study group was small. The follow up was of six months to one year only. Long term follows up and large numbers in the study group is needed to see the effect of LA reduction for restoration of sinus rhythm and thromboembolic complications. Very few patients reverted to sinus rhythm following LA reduction, hence definite conclusion could not be drawn regarding atrial fibrillation CONCLUSION Following MVR significant improvement in the NYHA functional class was noted in all the patients. The decrease in RVSP, LA size, and TR was noted in all the patients. The decrease noted in RVSP, TR and LA size were more in patients who underwent LA reduction. This decrease was noted up to six months to twelve months after follow up. Patients in AF with large LA, after reduction only 20% reverted to sinus rhythm. The plication technique is effective in reducing LA size, without much increase in cross clamp time, and there was no complication due to surgery. It seems reasonable to suggest that patients who undergo LA reduction along with MVR have significant improvement in clinical outcome and NYHA functional class with less thromboembolic complications during long term

follow up. Hence, LA reduction is advisable in patients with enlarged LA along with MVR. Identification of novel genetic loci for osteoporosis and/or rheumatoid arthritis using cFDR approach There are co-morbidity between osteoporosis (OP) and rheumatoid arthritis (RA). Some genetic risk factors have been identified for these two phenotypes respectively in previous research; however, they accounted for only a small portion of the underlying total genetic variances. Here, we sought to identify additional common genetic loci associated with OP and/or RA. The conditional false discovery rate (cFDR) approach allows detection of additional genetic factors (those respective ones as well as common pleiotropic ones) for the two associated phenotypes. We collected and analyzed summary statistics provided by large, multi-center GWAS studies of FNK (femoral neck) BMD (a major risk factor for osteoporosis) (n = 53,236) and RA (n = 80,799). The conditional quantile-quantile (Q-Q) plots can assess the enrichment of SNPs related to FNK BMD and RA, respectively. Furthermore, we identified shared loci between FNK BMD and RA using conjunction cFDR (ccFDR). We found strong enrichment of p-values in FNK BMD when conditional Q-Q was done on RA and vice versa. We identified 30 novel OP-RA associated pleiotropic loci that have not been reported in previous OP or RA GWAS, 18 of which located in the MHC (major histocompatibility complex) region previously reported to play an important role in immune system and bone health. We identified some specific novel polygenic factors for OP and RA respectively, and identified 30 novel OP-RA associated pleiotropic loci. These discovery findings may offer novel pathobiological insights, and suggest new targets and pathways for drug development in OP and RA patients. Introduction Osteoporosis (OP) is a major public health problem, affecting millions of people worldwide, particularly postmenopausal women. In the United States, the prevalence of OP is estimated to PLOS be greater than 14 million people in 2020 [1]. In China, the population with OP is expected to reach 212 million people by the year 2050 [2]. OP is characterized by low bone mass and micro-architectural deterioration of bone tissue, leading to increased bone fragility and increased susceptibility to fracture [3]. Bone mineral density (BMD) measurement is currently the standard major method for the diagnosis OP and fracture prediction. OP is a polygenetic disorder with a BMD heritability estimated between 50% and 80% [4]. Rheumatoid arthritis (RA) is an autoimmune inflammatory disease affecting approximately 0.5-1% of the world population. RA is characterized by symmetric joint swelling, tenderness and destruction of synovial joints, eventually leading to severe joint destruction and disability [5]. In US, RA affects 1.3 million adults [6]. In China, it is estimated that 0.32%-0.36% people suffer from RA [7]. It is widely recognized that RA is closely associated with both genetic and environmental factors. Twin studies showed that the heritability of RA is about 60% [8]. RA is frequently associated with OP, which may occur in 60-80% of RA patients [9]. In RA, bone is a target of inflammation, which leads to systemic OP, periarticular osteopenia and bone erosion, resulting in dramatically increased risks of fractures and related morbidity, mortality, and healthcare costs. Several cross-sectional studies have shown that a lower BMD was associated with increase risk in OP in patients with RA compared to matched controls [10,11]. OP is a multifactorial condition, especially in patients with RA. Many studies have demonstrated that OP are closely related to proinflammatory cytokines, such as TNF-a, IL-1, IL-6 and IL-17, which have been shown to be vital in bone resorption [12]. Genetic and environmental factors both contribute to RA-associated OP, and genetic background plays an important role. However, genetic risk factors have not been well defined in RA-associated OP. Research for genetic factors has been focused on screening for polymorphisms in inflammatory cytokine genes and OP candidate genes, such as VDR gene and OPG gene, to identify shared genes between OP and RA. Current findings accounted for only a small portion of the underlying total genetic variances for OP and RA. In addition, reported results from different studies were largely in disagreement [13][14][15][16]. Hence, it is necessary to systematically identify additional common genetic loci associated with OP and RA using a different and new approach. The conditional false discovery rate (cFDR) is a method recently developed by Andreassen et al. [17], and is powerful for genome-wide association studies (GWAS) analysis using summary statistics. Andreassen et al. has effectively applied and identified previously unsuspected common genetic risk loci for complex diseases, e.g., schizophrenia and cardiovascular disease traits [18], schizophrenia and bipolar disorder [17]. The principle of this method is to test variants for association with the principal phenotype conditional on different strengths of association with a second trait via combining the summary statistics from two different and independent GWAS of two related study traits [17]. We have applied the method and successfully identified some interesting shared genetic risk loci for common complex diseases, e.g., height and FNK BMD [19], Type 2 Diabetes and birth weight [20], Coronary artery disease and BMD [21]. To date, many GWAS for OP and RA have been performed separately and identified multiple genetic risk loci associated with them respectively. The cFDR method can efficiently increase the effective sample size for GWAS analysis of those pleiotropic loci based on existing GWAS data (summary statistics) of two related traits/diseases, hence no need to increase the original GWAS sample sizes with additional sample and data. In this study, we analyzed two large publically accessible GWAS datasets to identify common genetic loci for OP and RA using this new and powerful cFDR method. GWAS datasets We acquired two GWAS summary statistics from publically accessible datasets. The first dataset is the association analysis with RA coming from a GWAS meta-analysis of 80,799 subjects (European & Asian) carried out by Okada et al. [22], and we chose European-specific results with 58,284 subjects for this study, to ensure consistency of ethnicity with those used with the following study of OP. This was a meta-analysis that was obtained from 18 studies from Europe to assess European-specific association between RA and more than 8 million genotyped or imputed single-nucleotide polymorphisms (SNPs). The data included summary statistics, which provided the p-values showing association and effect of orientation for each variant after controlling for genomic inflation (due to potential population sub-structure) at both the individual study level and the meta-analysis. The 2nd dataset is the FNK BMD coming from a GWAS meta-analysis with 33 individual studies, which involved 32,735 European subjects and was performed by the Genetic Factors for OP (GEFOS) Consortium [23]. The dataset provided summary statistics for the associations of the FNK BMD and more than 10 million genotyped and imputed SNPs. The above two GWAS datasets are independent with no overlapping individuals. Data preparation Firstly, we annotated and mapped SNPs from the two GWAS studies to genes, and combined the summary statistics from the two GWAS studies for the 5,806,739 SNPs that were common in the two studies. Next, in order to generate SNPs being in approximate linkage equilibrium with each other, we applied a linkage disequilibrium (LD) based pruning method in Plink version 19.0 [24]. The pruning process of the dataset used the CEU HapMap 3 genotype and LD information. The pruning algorithm proceeded with a window of 20 SNPs, in which LD between each pair of SNPs was assessed, and if the LD was greater than 0.2 then one of the pair of SNPs was removed. The LD threshold of 0.2 was consistent with other using the cFDR method papers. For the pairs with the LD greater than 0.2, we adopted minor allele frequency (MAF) pruning method and removed the SNP of the smaller MAF. After this initial removal of SNPs, the window shifted 5 SNPs forward, and repeated the procedure until no pairs of SNPs being in high LD. Finally, there were 176,331 SNPs remained to be used in the analysis after pruning. In order to ensure that the genetic variance estimates from each SNP are not inflated owing to population structure, adjustment of GWAS results with genomic control is necessary. Because the authors from the two original GWAS datasets had previously applied genomic control rigorously, there was no need to apply this adjustment again in this study. Statistical analysis The concept of the cFDR extends from the unconditional false discovery rate (uFDR), which is defined as the probability of a false positive association for a set of variants in the single phenotype case [25]. Similarly, the cFDR applies this idea to the two related phenotype case and is characterized as the probability that a given SNP has a false positive association with the principal phenotype given that the p-values for both the principal and conditional phenotypes are less than or equal to the observed p-values. where the observed strength of association that a given SNP is associated with the principal phenotype is expressed as p i , and the observed significance level that the same SNP is associated with the conditional phenotype is expressed as p j . The H ðiÞ 0 represents the null hypothesis that a given SNP is not associated with the principal trait [19]. According to the steps outlined by Andreassen et al. [17], we computed the cFDR for each variant when FNK BMD is the principal phenotype conditioned on strength of association with RA (FNK BMD | RA) as well as the reverse (RA | FNK BMD). In order to confirm if the cFDR method leads to the enrichment of associated SNPs, we consecutively restricted the subset of SNPs that were tested based on the level of significance for the correlation of each variant with the conditional trait applying the following criteria for P j p j , such as P j 1 (all SNPs), P j 0.1, P j 0.01, P j 0.001. SNPs with the cFDR value smaller than 0.05 were regarded to be significantly related to the principal phenotype. In order to evaluate the enrichment of association in comparison with that expected under the null hypothesis, we drew conditional quantile-quantile (Q-Q) plots based on changing levels of significance in the conditional trait. The Q-Q plot shows a graph with the observed distribution of p-values plotted versus the expected distribution of p-values under the null. The yaxis represent the nominal p-values (-log 10 (p)) and the x-axis represent the empirical quantiles (-log 10 (q)). The degree of leftward shift from the expected null line intuitively shows pleiotropic enrichment. Conditional Manhattan plots were used to display the independent loci associated with RA, given their association with FNK BMD as well as the reverse. All the SNPs after pruning within an LD block and their chromosomal location in the genome are shown in these plots. When a-log 10 (p) value is greater than 1.3, the variant was considered to be significantly associated with the principal phenotype considering the association of that variant with the conditional phenotype. To identify loci that are associated with both FNK BMD and RA, we computed the conjunction cFDR (ccFDR) value after computing the cFDR(FNK BMD | RA) as well as the reverse cFDR(RA | FNK BMD). The ccFDR is characterized as the probability that a given SNP has a false positive association with both the principal and conditional phenotypes. The maximum cFDR value of the two cFDR values is selected as the ccFDR [17]. All SNPs with a ccFDR value smaller than 0.05 were determined to be significantly associated with both traits. In order to show the locations of pleiotropic genetic variants, we presented a conjunction Manhattan plot. Functional term enrichment analysis We performed functional term enrichment analyses using the GO terms database [26] and evaluated the potential functions of the trait associated loci identified by cFDR and pleiotropic loci identified by ccFDR. We then classified trait-associated loci according to their known biological processes and molecular functions. This bioinformatic analysis is intended to partially assess the validity and functional relevance of our findings by identifying gene sets that are significantly related to bone metabolism and/or RA processes. Assessment of pleiotropic enrichment Conditional Q-Q plots are used to visualize the distribution of p-values and assess the pleiotropic enrichment. The distribution of p-values follows the null distribution diagonal line under the null hypothesis. The enrichment of genetic association would appear to deflect leftwards from the null line

because the association of principal phenotype depends on the more stringent association criteria of the conditional phenotype. The larger the interval between the conditions Q-Q plots, the greater the degree of pleiotropic-effect of the genes shared between the two traits. The conditional Q-Q plot of the FNK BMD giving nominal p-values associated with RA ( Fig 1A) reveals some enrichment at the different significance thresholds of RA. The presence of leftward shift indicates that the number of true associations for a given RA p-value increase when the analysis is limited to include SNPs of more significant associations. From the conditional Q-Q plot for RA given FNK BMD (Fig 1B), we can observe a larger separation between the different curves, which indicate that there may be a stronger enrichment for RA (given FNK BMD) than that for FNK BMD (given RA). FNK BMD loci identified with cFDR We identified a total of 88 significant SNPs (cFDR < 0.05) for FNK BMD variation associated with RA (Fig 2A), which were located on 16 different chromosomes. Of the 88 SNPs, 53 had pvalues smaller than 1x10 -5 and 18 reached genome-wide significance at 5x10 -8 in the original meta-analysis for FNK BMD [23]. There are four SNPs among the eighteen SNPs-rs7554551 [27], rs6710518 [28], rs1038304 [29], rs228768 [30], that were previously reported/identified to be associated with BMD or OP, and the other fourteen were not found reported/identified in earlier studies. Furthermore, these 88 variants identified to be associated with FNK BMD are enriched in several skeletal metabolism related functional terms (Table 1), such as "positive regulation of bone mineralization", "negative regulation of ossification", "osteoblast development", "positive regulation of chondrocyte differentiation", and "negative regulation of canonical Wnt signaling pathway". RA gene loci identified with cFDR We identified a total of 778 significant SNPs (cFDR < 0.05) for RA variation associated with FNK BMD (Fig 2B), which were located on 22 different chromosomes. Of the 778 SNPs identified, 567 had p-values smaller than 1x10 -5 and 437 reached genome-wide significance at 5x10 -8 in the original meta-analysis for RA [22]. In order to carry out the functional term enrichment analysis for these results, we chose 633 variants with cFDR < 0.01 for analysis. The results shown a large number of RA loci were enriched in several immune system related functional terms (Table 1), such as "immune response", "antigen processing and presentation", "T cell costimulation", "positive/negative regulation of T cell proliferation", "negative regulation of tumor necrosis factor production", "negative regulation of inflammatory response", "innate immune response", and "regulation of immune response". Pleiotropic gene loci for both FNK BMD and RA We applied the ccFDR value to identify genetic loci associated with both FNK BMD and RA. The ccFDR is characterized as the probability that a given SNP has a false positive association with both the principal and conditional phenotypes. Furthermore, we constructed the ccFDR Manhattan plot (Fig 2C) to graphically show the distribution of the pleiotropic loci identified with the two traits. We identified 30 independent pleiotropic loci significantly (ccFDR < 0.05) associated with both traits ( Table 2). 18 of the 30 variants were located within the region of the major histocompatibility complex (MHC), which is a genomic region located at 6p21.3 and is essential in the control of the immune system [31], and is previously confirmed to be associated with RA [32] and OP [33]. Of the 18 SNPs, 13 SNPs were located within the class III region of the MHC-rs9378164 (LY6G5C), rs1266071 (BAT5), rs2734335 (C2), rs401775 (SKIV2L), rs6909427, rs547261, rs9368716, rs1265762, rs2073046 (C6orf10), rs3115576, rs9268055, rs2395114 and rs4713518 (NOTCH4 and C6orf10), and other 5 SNPs were located within the class II region of the MHC-rs35571244 and rs35599935 (PPP1R2P1 and HLA-DMB), rs34043227 (HLA-DQB2 and HLA-DOB), rs3101944 (HLA-DMB and HLA-DMA), and rs1367727 (HLA-DMA and BRD2). Twelve other variants of the 30 identified pleiotropic variants are located on 6 different chromosomes: seven variants ((rs17016749 (INPP4B), rs2880389 (BOLL and PLCL1), rs1318867 and rs11684176 (PLCL1), rs11066320 (PTPN11), rs1109241 (ETS1 and FLI1), and rs9469623 (LOC100132252 and GRM4)) are located at or close to the genes that are related to RA and other immune associated traits or OP and other bone related traits [22,[34][35][36][37][38], and five variants ((rs13 413470 (RFTN2), rs7340470 (RFTN2 and MARS2), rs700653 (BOLL), rs10947432 (C6orf125), and rs2099102 (EHD2)) are novel and not related to any of the two diseases or related traits after a thorough literature search (OMIM and PubMed). Discussion In this study, we discovered 88 FNK BMD susceptibility loci, which included 18 loci reached genome-wide significance at 5x10 -8 in the original meta-analysis for FNK BMD [23]. And we also identified 778 RA susceptibility loci, which included 567 loci reached genome-wide significance at 5x10 -8 in the original meta-analysis for RA [22]. Most importantly, we reported 30 new OP-RA associated pleiotropic loci that have not been reported in previous OP or RA GWAS. Our study represents the first systematic effort to explore the genetic basis of OP and RA in a large sample set. We identified the pleiotropic effects of variants associated with OP and RA by combining the summary statistics from two different GWAS meta-analyses. Compared with the existing standard single phenotype analysis, synchronous analysis for multiple related traits not only increase discovery of trait-associated variants under original datasets for individual traits, but also provide a new insight to further detect the common genetic mechanisms between associated phenotypes. Gene ontology enrichment analysis was carried out using David [26], and "P-value<0.05" suggests the term is significantly more enriched than random chance. We carried out the pleiotropic analysis using a cFDR approach, which was recently developed by Andreassen et al. [17]. The basic idea of cFDR is that variants are more likely to exert a true effect when the variants have significant effects in both traits. In the original simulation of the cFDR approach, the authors demonstrated that cFDR approach resulted in 15-20 times' increase of the number of non-null SNPs with a local FDR smaller than 0.05, when compared with the uFDR. It's well known that the traditional meta-analysis method can also increase statistical power, but it can only detect loci that have largely or dominantly the same direction of allelic effects on both traits. In contrast, cFDR method can detect loci regardless of their effect directions [39]. In the following, we highlight some of the salient discoveries from this study for their potential significance to OP and/or RA. MHC Among the 30 SNPs associated with both FNK BMD and RA, 18 were located within MHC region, which is the most vital genomic regions associated with RA. Those 18 SNPs are located Osteoporosis and/or rheumatoid arthritis at or near the following genes that have been reported to be related to RA: C6orf10 [40], NOTCH4 [41], HLA-DQB2 [42], HLA-DMB [43], HLA-DMA [43], and BRD2 [44]. Although we do not find any report about association between these gene and OP or related traits, some other MHC genes have been identified to be associated with OP in previous research [33]. In addition, it was estimated that approximately 40% of the expressed genes in MHC have immune system function [45]. There is increasing evidence of an association between the immune system and bone [46]. The bone remodeling could be influenced by proinflammatory cytokines (eg., IL-1, IL-6, TNF-a, M-CSF, PGE2) primarily through regulating the osteoprotegerin (OPG) / receptor activator of nuclear factor-kB (RANK) / RANK ligand (RANKL) [12]. PLCL1 The two pleiotropic loci (rs1318867, rs11684176) were located in the intron of PLCL1 (phospholipase C like 1), an OP candidate gene, which was found to be associated with variation in hip bone size [34]. As for the relationship between PLCL1 and RA, at present, there is no study reported. However, variants of the PLCL1 gene had been proven to be associated with other autoimmune diseases, such as Crohn's disease (CD) [47], systemic lupus erythematosus (SLE) [35]. Clinically, multiple autoimmune diseases (ADs) can be observed in cluster [48], including families with RA, suggesting some degree of common genetic susceptibility. PTPN11 The SNP rs11066320 is located in the intronic region of the PTPN11 (protein tyrosine phosphatase, non-receptor type 11) gene, which encodes Src homology region 2 (SH2)-containing protein tyrosine phosphatase 2 (SHP-2). SHP-2 belongs to the protein tyrosine phosphatase family, which have been involved in regulating intracellular signal transduction initiated via multiple different growth factors and cytokine receptors [49]. A recent GWAS reported that a RA risk locus (dbSNP ID = rs10774624) is related to the PTPN11 gene, indicating a role for PTPN11 in RA [22]. Additionally, the PTPN11 gene is significantly overexpressed in RA fibroblast-like synoviocytes (FLS) compared with osteoarthritis (OA) FLS, suggesting a novel role for SHP-2 in promoting RA FLS invasiveness [50]. On the other hand, SHP-2 acts as a vital role in the RANK/NFATc1 signaling pathway, and regulates osteoclastogenesis and bone mineral homeostasis [51]. Furthermore, SHP-2 has been shown to be important in controlling osteoblast differentiation, proliferation and metabolism [52]. ETS1 The SNP rs1109241 (11p11.12) is located close to the ETS1 (ETS proto-oncogene 1, transcription factor) and FLI1 (Fli-1 proto-oncogene) gene, which have been reported to be associated with RA susceptibility [53]. ETS-1 is part of the ETS family of transcription factors, primarily expressed in lymphoid cells, acts as relevant roles in the lymphocyte development, apoptosis and inflammation and controls the expression of abundant immune-related genes [54]. Several GWAS demonstrated that ETS1 is associated with RA in Europeans [22], and Asians [7]. Furthermore, Ets-1 is also expressed in pre-osteoblast and osteoblast cells, and studies have revealed that Ets-1 exerts profound effect on osteoblast differentiation and bone development via various mechanisms [55]. INPP4B The SNP rs17016749 (4q31.21) is located in the intronic region of the INPP4B (inositol polyphos-phate-4-phosphatase type II B) gene, which encodes the inositol polyphosphate 4-phosphatase type II, one of the enzymes involved in phosphatidylinositol signaling pathways [37]. Ferron et al. revealed that INPP4B acts as a major regulator of osteoclast differentiation by modulating calcium signaling and NFATc1 pathway. They also demonstrate that INPP4B in humans is a novel determinant of OP susceptibility locus via genetic analysis [56]. In summary, our study demonstrated that there is significant pleiotropy between OP and RA by assessing pleiotropic effects in the framework of a conditional analysis. We identified several novel pleiotropic loci for OP and RA. Our results bring further insight into the common genetic influences of OP and RA. Hence, in order to confirm and explore the generality of the results identified in this study, other ethnic populations and animal or cell experiments are needed in future studies. MOR Is Not Enough: Identification of Novel mu-Opioid Receptor Interacting Proteins Using Traditional and Modified Membrane Yeast Two-Hybrid Screens The mu-opioid receptor (MOR) is the G-protein coupled receptor primarily responsible for mediating the analgesic and rewarding properties of opioid agonist drugs such as morphine, fentanyl, and heroin. We have utilized a combination of traditional and modified membrane yeast two-hybrid screening methods to identify a cohort of novel MOR interacting proteins (MORIPs). The interaction between the MOR and a subset of MORIPs was validated in pulldown, co-immunoprecipitation, and co-localization studies using HEK293 cells stably expressing the MOR as well as rodent brain. Additionally, a subset of MORIPs was found capable of interaction with the delta and kappa opioid receptors, suggesting that they may represent general opioid receptor interacting proteins (ORIPS). Expression of several MORIPs was altered in specific mouse brain regions after chronic treatment with morphine, suggesting that these proteins may play a role in response to opioid agonist drugs. Based on the known function of these newly identified MORIPs, the interactions forming the MOR signalplex are hypothesized to be important for MOR signaling and intracellular trafficking. Understanding the molecular complexity of MOR/MORIP interactions provides a conceptual framework for defining the cellular mechanisms of MOR signaling in brain and may be critical for determining the physiological basis of opioid tolerance and addiction. Introduction Opioid agonist drugs are clinically important because they are potent analgesics. However, chronic exposure to opioid drugs causes profound changes in the brain, which may lead to opioid dependence. The analgesic and addictive properties of most clinically relevant opioid agonist drugs are mediated primarily via activation of mu-opioid receptors (MORs). The central role of MOR in mediating the effects of opioid agonist drugs was established using MOR knockout (KO) mice. MOR KO mice display significantly reduced sensitivity to

both the analgesic and rewarding effects of opioids [1]. Regulation of MORs, like most G-protein-coupled receptors (GPCRs), occurs via multiple mechanisms including receptor desensitization, internalization, degradation, and recycling [2]. A number of studies have shown that MOR desensitization and receptor trafficking can increase the rewarding properties of opioid drugs, while reducing the development of opioid tolerance and addiction-like behaviors [3,4,5,6,7,8,9,10,11]. However, the spe-cific molecular mechanisms that regulate these processes are largely unknown. Elucidating the mechanisms that regulate MOR signaling and trafficking is critical for determining the cellular response to opioid agonist drugs and for opening new avenues of investigation into the pharmacotherapy of pain management. A fundamental principle that has emerged from decades of cell signaling research is that signaling molecules, including GPCRs, are assembled into macromolecular entities termed signaling complexes or signalplexes [12,13]. It is now well established that receptor-protein interactions govern the structural and functional organization of GPCR-containing signalplexes [14,15,16,17]. To date, more than 20 proteins that interact directly with the MOR (MORIPs; mu-opioid receptor interacting proteins) have been identified. These interacting proteins have been shown to play a role in regulation of MOR trafficking, subcellular localization, and signaling [18]. Additionally, activation of the MOR can affect the function of some of its interacting partners. For example, we have shown that morphine promotes the interaction between the MOR and WLS (Wntless/GPR177; a protein required for Wnt protein secretion), and this interaction serves to inhibit Wnt protein secretion from transfected mammalian cells [19,20]. To better understand the potential role of MORIPs in the MOR life cycle, we have initiated traditional and modified yeast two-hybrid (Y2H) studies designed to identify novel constituents of the MOR signalplex. Previous interaction screens for MORIPs have primarily utilized the third intracellular loop (IL3) or the Cterminal tail (C-tail) of the MOR as bait [18]. Here we have utilized the second intracellular loop as well as the entire MOR to screen human brain cDNA libraries in order to expand the growing list of MORIPs. Using these approaches, we have identified ten novel MOR binding partners, validated their interaction with the MOR, and examined the expression of three of these proteins in the brains of morphine-treated mice. In addition, we investigated whether two newly identified MORIPS, SIAH1 and SIAH2, are involved in ubiquitination or proteolysis of MOR. Further functional characterization of MORIPs will serve to heighten our understanding of the mechanisms regulating MOR-mediated signaling and may help elucidate the underlying molecular basis of cellular response to opioid agonist drugs. Traditional and MYTH Two-Hybrid Screens The traditional Y2H screening method involves the reconstitution of the GAL4 transcription factor through the interaction of a bait protein fused to the GAL4 DNA binding domain and a prey protein (from a fetal brain cDNA library) fused to the transcriptional activating domain of GAL4 [21]. This method is biased for cytosolic and nuclear proteins, as the protein complex must be imported into the nucleus to activate transcription. Therefore, only the cytosolic portion of integral membrane proteins is usually employed in this type of screen. Traditional Y2H screens in this study were performed as previously described [22,23] using the second intracellular loop (IL2; amino acids 166-187) of the MOR as bait to screen a fetal human brain cDNA library. The MOR-IL2 was cloned into the yeast GAL4 DNA binding domain expression vector pAS2-1 (Clontech, Palo Alto, CA), while the human fetal brain cDNA library was provided in the GAL4 activation domain vector pACT2 (Clonetech). Bait and prey plasmids were successively transformed into yeast strain MaV103 [22]. Transformation of yeast with the fetal human brain library produced ,2610 6 transformants/mg of DNA on quadruple dropout plates (-Leu/2Trp/2His/2Ura; Clonetech) containing 3-amino-1,2,4-triazol (3AT). Interactions were assayed for bgalactosidase (b-gal) activity via the nitrocellulose lift method [22]. cDNAs were extracted from yeast colonies, sequenced, and subjected to Basic Local Alignment Search Tool (BLAST) analysis to determine their identities. To identify additional MOR interacting proteins (MORIPs), a modified split-ubiquitin membrane yeast two-hybrid (MYTH) screen was performed as previously described [24]. The MYTH system uses the split-ubiquitin method, in which the reconstitution of ubiquitin is mediated by a specific protein-protein interaction. Ubiquitin-specific proteases cleave at the C-terminus of ubiquitin, which releases a transcription factor that can translocate to the nucleus and activate transcription of a reporter gene [25]. The unique advantage of MYTH is that full-length integral membrane proteins can be used as bait and are amenable to protein-protein interaction analyses in their natural membrane environment [26,27]. For this study, full-length human MOR cDNA in the bait vector pCCW-STE (Dualsystems Biotech AG, Switzerland) and a human fetal brain library in the prey vector pPR3-N (Dualsystems) were sequentially transformed into S. cerevisiae reporter strain THY.AP4. Transformation of yeast with the human brain library yielded 6610 6 transformants/mg DNA on quadruple drop out plates (2Trp/2Leu/2His/2Ade; Clonetech) containing 3AT. Fifty transformants were positive for b-gal activity. These colonies were picked and their cDNAs extracted, sequenced and subjected to BLAST analysis. From this screen we identified four novel MORIPs ( Table 1) that were subjected to further biochemical analysis. To map sites of interaction between the MOR and the newly identified MORIPs, each MOR intracellular loop (IL) was tested for interaction with individual MORIPs using the traditional Y2H method. MOR-IL domains (IL1, amino acids 97-102; IL2, amino acids 166-187; IL3, amino acids 259-282; and C-tail residues 361-420) were separately ligated into pAS2-1 and assayed for interaction with candidate MORIP cDNA clones in pACT2. Bait and prey plasmids were simultaneously co-transformed into S. cerevisiae strain MaV103 and interactions assayed for b-gal activity as described above. Treatment times were based on previous experiments by Hislop et al. [32]. Animal Care and Drug Treatment All mouse studies were performed in the research laboratories at the Department of Veterans Affairs Medical Center in Coatesville, PA, and were approved by the Institutional Animal Care and Use Committee at both DVAMC Coatesville and the University of Pennsylvania. C57BL/6J mice were bred in-house, maintained on a 14 hr/10 hr light/dark schedule and had food and water available ad libitum throughout the course of the experiment. Female mice, 8-9 weeks of age, were implanted sub-cutaneously with a 25 mg morphine (n = 5) or placebo (n = 4) pellet. Morphine and placebo pellets were obtained from the NIDA Drug Supply Program. Mice were euthanized by cervical dislocation under carbon dioxide anesthesia four days (96 hr) after pellet implantation. Brains were harvested and dissected on ice under 56 magnification into 6 regions including prefrontal cortex, nucleus accumbens, dorsal striatum, midbrain, hippocampus, and cerebellum. The atlas of Paxinos and Watson [33] was used as a guide. Specimens were frozen on dry ice and stored at 280uC. Tissue Preparation and Protein Analysis Frozen brain tissue obtained from drug-or sham-treated mice was utilized for protein expression analysis. Fresh whole-brain tissue from untreated mice was obtained for co-IP experiments. All tissue was suspended in lysis buffer (50 mM Tris-HCl, 1 mM EDTA, 150 mM NaCl, 1% NP40, 0.25% deoxycholate, 5 mM NaF, 2 mM Na 3 VO 4 ) containing protease inhibitors (cOmplete MINI EDTA free, Roche), homogenized using a microcentrifuge pestle for 2 minutes, sonicated using a probe sonicator, then centrifuged at 13,000 RPM to remove cellular debris. For protein expression analysis, samples were analyzed by electrophoresis on 10% SDS-containing polyacrylamide gels, followed by Western blotting. In addition to antibodies utilized for co-IP, guinea pig anti-MOR (GP10106, Neuromics, Edina, MN) and rabbit anti-DOK4 (SAB1300112, Sigma-Aldrich) antibodies were also used to probe brain lysates. For quantitative analysis of MORIP expression in drug treatment studies, lysates from sham and morphine-treated mice (20 mg/brain region) were analyzed in quadruplicate. To rule out transfer artifacts, samples from each brain region were loaded in a different order on each of the four blots. Blots were scanned using a back-lit scanner and quantification was performed using IMAGEJ software [34]. Expression was normalized to total protein (as measured by Ponceau stain), averaged between replicates, and subjected to a two-tailed Students t-test. Identification of Novel MORIPs To identify novel mu-opioid receptor interacting proteins (MORIPs), we used the second intracellular loop of the MOR as bait to screen a fetal human brain cDNA library. In control experiments, the MOR-IL2 construct used in these studies did not autoactivate b-galactosidase expression (data not shown). We screened 2610 6 colonies and isolated 30 positive (b-gal+) clones representing eight distinct human proteins (Table 1). Sequence analysis identified several clones encoding components involved in the ubiquitin pathway and in signal transduction. These included COP9 subunit 5 (CSN5), a protein proposed to regulate exosomal protein sorting in both a deubiquitinating activity-dependent andindependent manner [35], ancient ubiquitous protein 1 (AUP1), a protein that promotes ubiquitination of misfolded proteins [36], and seven in absentia homologs 1 and 2 (SIAH1 and SIAH2), known RING Finger E3 ubiquitin ligases [37]. Four putative scaffolding proteins, docking protein 4 (DOK4), docking protein 5 (DOK5), Zyxin, and Ran binding protein 9 (RanBP9) were also identified as candidate MORIPs (Table 1). To map the binding site of the MORIPs on the MOR, we used the traditional yeast two-hybrid system to interrogate interaction of individual MORIPs with each of the MOR intracellular loops and the C-terminal cytoplasmic tail. Each of the MORIPs we identified interacted with IL2 of the MOR ( Fig. 1 and Table 2). Two of the MORIPs, CSN5 and RanBP9, also interacted with MOR-IL1, while AUP1, CSN5, DOK4 and RanBP9 were found to interact with the C-tail of the receptor. The binding site for Cx37 was not tested in this assay. These results suggest that the MORIPs we identified interact with specific domains of the MOR in the yeast system, and that the second intracellular loop of the receptor is an apparent hot-spot for protein-protein interactions. Validation of Direct MOR/MORIP Interactions To further validate direct interaction between newly identified MORIPs and the MOR, we tested the ability of selected MORIP-GST fusion proteins to associate with S-tagged MOR-IL2 in a pull-down assay. The results are shown in Fig. 2 and summarized in Table 2. Western blots containing lysates from bacteria expressing an S-tagged MOR-IL2 cDNA produced an immunoreactive band of ,13 kDa when probed with anti-S-tag antibodies. This band corresponds to the expected size of the MOR-IL2 encoded by the cDNA construct. The same band was detected by pull-down after the bacterial lysate was incubated with the CSN5-GST, AUP1-GST, or DOK4-GST fusion proteins, but not when the lysate was absorbed onto beads alone or GST-coated beads. We also utilized the pull-down assay to test the ability of RanBP9, SIAH1, and Zyxin to associate with a MOR-IL2-GST fusion protein. As shown in Fig. 2, S-tagged RanBP9 cDNA produced an immunoreactive band of ,34 kDa, S-tagged SIAH1 cDNA produced a band of ,32 kDa, and S-tagged Zyxin was detected at ,85 kDa when probed with anti-S-tag antibody. These bands correspond in size to the predicted sizes of the protein fragments encoded by the respective cDNA constructs, and were not detected when the lysates were absorbed onto beads alone or GST-coated beads. Interaction of MORIPs with MOR, DOR, and KOR in Cultured Mammalian Cells The interaction between full-length MOR and full-length MORIPs was verified using co-IP experiments. To demonstrate an interaction in cell culture, we tested the ability of an anti-FLAG antibody to co-IP MORIPs from lysates prepared from HEK-MOR cells that stably express FLAG-tagged MORs and have been transiently transfected with myc-tagged MORIPs. Probing Western blots with anti-myc antibodies revealed immunoreactive bands of the expected sizes in lysate lanes (L) prepared from HEK-MOR cells transiently transfected with myc-tagged AUP1, CSN5, Cx37, DOK4, SIAH1, SIAH2, VAPA, or WLS cDNAs (Fig. 3A). Bands migrating at similar molecular weights were detected in the co-IP lanes, but not in negative controls. Independent Y2H and co-IP experiments have recently confirmed an interaction between RanBP9 and the MOR [47]. Taken together, these results support the view that the MORIPs identified in our Y2H screens interact with full length MOR in the context of cultured mammalian cells (summarized in Table 2). Our Y2H and pulldown data indicate that each of the MORIPs identified in our screens interacts with the MOR-IL2 (Fig. 1). Sequence comparisons indicate that the IL2 of the MOR, DOR and KOR exhibit a high degree of amino acid sequence similarity (Fig. 3A). Among the opioid receptor proteins, IL2 exhibits 85-90% amino acid sequence identity, with the greatest conservation within the

N-terminal portion of the loop. Because of this homology, we used co-IPs to determine whether any of the MORIPs also interacted with other opioid receptor family members. To do this, we tested the ability of anti-FLAG antibodies to co-IP MORIPs from lysates prepared from HEK-DOR and HEK-KOR cells (stably expressing FLAG-tagged DORs and KORs, respectively) and separately transfected with each of the myc-tagged MORIPs. As shown in Fig. 3B and summarized in Table 2, AUP1, Cx37, SIAH1, SIAH2, VAPA, and WLS were able to interact with both DOR and KOR. However, neither CSN5 nor DOK4 showed interaction with DOR or KOR, despite the fact that the proteins were expressed in transfected HEK-DOR and HEK-KOR cells (Fig. 3B). Together, these results suggest that many of the MORIPs we identified are also capable of interaction with the DOR and KOR, at least within the context of transfected mammalian cells. Thus, it may be more precise to describe these interactors as ORIPs (opioid receptor interacting proteins), rather than MORIPs. The interaction of these proteins with each of the opioid receptor family members will depend in large part on whether an opioid receptor family member and a particular opioid receptor binding protein are co-expressed within the same cell (and cellular compartment) within the nervous system. Interaction of MOR and MORIPs in Rodent Brain To determine whether the interaction between the MOR and MORIPs may occur in vivo, we analyzed the expression of MOR MORIPs isolated either in traditional or MYTH screens were tested for interaction with each of the intracellular loops (IL1, 2, or 3) or carboxyl-terminus (C-tail) of the MOR in a directed Y2H assay. 2 GST PD: Full-length MORIPs were tested for interaction with the MOR-IL2 in a GST pulldown assay. and several MORIPs in various mouse brain regions. As shown in Fig. 4A, the MOR and its interacting partners DOK4, SIAH1, VAPA, and WLS were each expressed in cerebellum, hippocampus, nucleus accumbens, midbrain, prefrontal cortex, and dorsal striatum. Co-expression of the MOR and each of the MORIPs in all brain regions tested is consistent with the idea that the MOR and its binding partners have the potential to interact in multiple brain regions. To verify interaction, the MOR was immunoprecipitated from a whole mouse brain lysate, and immunocomplexes probed for the presence of several MORIPs. As shown in Fig. 4B, SIAH1, VAPA, and WLS were able to co-IP with MOR but not with rabbit-IgG alone, suggesting that these MORIPs are capable of interacting with MOR in mouse brain. DOK4 was not tested due to crossreactivity of this antibody with rabbit IgG, while several other proteins were difficult to analyze due to the poor quality of available antibodies. The Effect of Morphine on MORIP Expression To determine whether chronic morphine exposure alters the protein levels of MORIPs, we performed Western blot analysis of selected MORIPs (DOK4, SIAH1, and WLS) in brain regions thought to be involved in response to opioid drug exposure. Mice were implanted subcutaneously with either saline (sham n = 4) or morphine (n = 5) pellets for 96 hours. As shown in Fig. 5, no significant changes were detected in any of the MORIPs tested in cerebellum, nucleus accumbens, or prefrontal cortex. However, in response to morphine exposure, DOK4 levels were significantly decreased by 53% in hippocampus (p = 0.01) and 33% in midbrain (p = 0.02). In morphine-treated animals, SIAH1 expression was reduced by 46% in hippocampus (p = 0.005). WLS levels were significantly decreased by 44% in midbrain (p = 0.01) and 30% in striatum (p = 0.03) following drug exposure. Together these data suggest that morphine treatment may lead to significant alterations in MORIP protein expression in various regions of mouse brain. SIAH 1 and SIAH2 do not Affect MOR Ubiquitination MOR is known to be ubiquitinated on the first intracellular loop, which affects sub-endosomal localization of the receptor The MOR was immunoprecipitated from whole mouse brain lysates using rabbit anti-MOR antibodies. Immunocomplexes were probed for the presence of SIAH1, VAPA, or WLS using MORIP-specific antibodies. Lysate lanes (L) contain 5% of the total protein prepared from whole mouse brain lysate compared to the mock (M, rabbit IGG) and immunoprecipitation (IP) lanes. doi:10.1371/journal.pone.0067608.g004 mu-Opioid Receptor Protein Interaction Screen PLOS ONE | www.plosone.org [32,48]. SIAH1 and SIAH2, two MORIPs identified in the current study, are RING finger E3 ubiquitin ligases that have been shown to ubiquitinate a variety of cellular proteins, including several surface receptors, and target them for degradation [49,50,51,52,53,54,55,56]. We, therefore, investigated whether SIAH1 and SIAH2 contribute to ubiquitination or degradation of the mu-opioid receptor. While one previous study has suggested a role of the proteasome in agonist-induced down-regulation and basal turnover of mu and delta opioid receptors [57], it is generally believed that, like other GPCRs, MOR degradation is mediated primarily via the lysosomal degradation pathway [32,58]. To confirm that this is the preferred mechanism of MOR proteolysis in the HEK-MOR cells used in our studies, we analyzed the effects of chloroquine (a lysosomal blocker) and MG132 (a proteosomal inhibitor) on Figure 5. MORIP expression in brain regions of morphine-treated mice. Mice were treated for 96 hours with either morphine-containing (n = 5) or placebo (n = 4) pellets. Animals were sacrificed, brain regions dissected, and Western blots of selected MORIPs probed with MORIP-specific antibodies. Each panel contains a representative blot for a MORIP in the specified brain region (n = 4 blots/MORIP/brain region). Total protein was quantified by Ponceau stain of the blot prior to antibody probing. Bar graphs represent the average pixel density (as determined by imageJ) of four blots for each brain region normalized to total protein and placebo treatment. Data was analyzed using a two-sided Student's t-test. Error is expressed as standard error of the mean. * indicates a statistically significant difference (p,.05) between sham and morphine treatment. As shown in Fig. 6A, blocking the lysosomal degradation pathway with 200 mM chloroquine for 4.5 or 9 hours led to an increase in MOR protein levels, whereas blocking the proteasomal degradation pathway with 30 mM MG132 for similar times led to a reduction in MOR levels. These results are consistent with previous reports that MOR proteolysis occurs predominantly via the lysosomal degradation pathway [32,58]. Ubiquitination of MOR species was analyzed by immunoprecipitating equal amounts (normalized to MOR levels in lysates) of MOR from cell lysates and probing for ubiquitin. It should be noted that with this protocol, it is possible that the ubiquitinated proteins detected may also include MORIPS that co-immunoprecipitate with MOR. Inhibition of lysosomal degradation with 200 mM chloroquine for 9 hours did not lead to an increase in ubiquitinated MOR (Fig. 6B). These results indicate that ubiquitination may not be required for lysosomal degradation of MOR. In contrast, we observed an increase in MOR ubiquitination after 9 hours of proteosomal inhibition with 30 mM MG132 (Fig. 6B). Because the proteasome is not required for basal turnover of MOR, we hypothesize that this increase may represent an indirect effect of stabilizing other proteins involved in the ubiquitination of MOR or an increase in the ubiquitination of associated MORIPS. DADLE, a mu and delta opioid receptor agonist that internalizes and promotes ubiquitination of MOR [32], was used to determine whether ligand-induced receptor degradation in HEK-MOR cells is regulated via the proteosomal or lysosomal pathway. In HEK-MOR cells, 6 hours of DADLE treatment lead to a mean decrease in MOR protein levels of 87% (Fig. 6C). In chloroloquine-or MG132-treated cells, DADLE induced a 53% or 48% reduction in MOR protein levels, respectively compared to cells that were not treated with inhibitor. In contrast to previous studies that implicate either the proteosome or the lysosome in the degradation of MOR [32,57], our results indicate that in HEK-MOR cells both pathways may contribute to agonist-induced MOR degradation. Because proteasome blockers increase the ubiquitination state of MOR and can partially rescue DADLE-induced degradation of the receptor, we asked whether these processes involve the function of either SIAH1 or SIAH2. SIAH1 and SIAH2 are E3 ubiquitin ligases that contain an N-terminal catalytic RING finger domain and a C-terminal substrate-binding domain. GST pulldown assays with the second intracellular loop (IL2) of MOR show that the MOR-IL2/SIAH1 interaction is limited to the N-terminal half of the substrate-binding domain of SIAH1 (amino acids 91-157; Fig. 7A). Based on this mapping study, we generated SIAH1 and SIAH2 mammalian expression constructs lacking their respective RING domains. These constructs should maintain an interaction with MOR while losing ubiquitin ligase capabilities. As shown in Fig. 7B, wild-type and truncated versions of SIAH1 and SIAH2 were able to interact with the MOR as evidenced by their ability to co-immunoprecipitate with the receptor. It is known that wild-type SIAH proteins autoubiquitinate [59] leading to low levels of detectable protein. Therefore, the cells used for this experiment were treated for 6 hours with MG132 to prevent degradation of the transfected SIAH1 and SIAH2 proteins. We next analyzed the effect of expressing wild-type or truncated SIAH1 and SIAH2 proteins on MOR levels in untreated and DADLE-stimulated HEK-MOR cells. Overexpression of wildtype or truncated SIAH proteins did not significantly alter steadystate MOR expression levels or DADLE-induced degradation of MOR (Fig. 7C). These results suggest that neither SIAH1 nor SIAH2 are involved in regulating MOR protein levels. Data from this study and that of Hislop et al. [32] indicate that the ubiquitination state of the MOR is independent of protein levels and is not required or sufficient for receptor degradation. However, it is possible that SIAH proteins may serve to ubiquitinate the MOR as a requisite for their proper intracellular trafficking. To assess whether SIAH proteins may play a role in MOR ubiquitination, we first analyzed the effect of wild-type or truncated SIAH constructs on steady-state MOR ubiquitination. Results from this analysis indicate that neither the wild-type nor the truncated forms of SIAH1 or SIAH2 had any noticeable effect on the steady-state ubiquitination of MOR (Fig. 7D). Treatment of cells with 10 mM DADLE for 30 minutes [32] or 30 mM MG132 for 9 hours (Fig. 6C) leads to an elevation in MOR (and/or MORIP) ubiquitination. To determine whether SIAH1 or SIAH2 are responsible for this increase in ubiquitination, we assessed the effects of overexpressing truncated SIAH constructs on MOR ubiquitination in DADLE or MG132-treated HEK-MOR cells. As shown in Figs. 7E and 7F, expression of truncated SIAH1 or truncated SIAH2 did not block ubiquitination induced by MG132 or DADLE. Taken together, our results suggest that neither SIAH1 nor SIAH2 plays a direct role either in steady-state or agonist-induced degradation or ubiquitination of the MOR. Discussion Using a combination of traditional Y2H and modified MYTH screening methods, we identified a novel cohort of mu-opioid receptor interacting proteins. In contrast to previous Y2H-based studies which employed either the IL3 or the C-tail of the MOR to discover potential binding partners, we used the MOR-IL2 as bait to screen a brain cDNA library. With this approach, we identified eight MORIPs. One of these interactors. RanBP9, has recently been confirmed as a MOR binding protein by an independent group [47]. Employing the entire MOR as bait in a MYTH screen, we identified four additional MORIPs, one of which (WLS) we have previously characterized in some detail [19,20,60,61]. In contrast to the MORIPs identified in the traditional Y2H screen, each of the MORIPs isolated in the MYTH screen represents an integral membrane protein. Because MYTH screens utilize the entire coding region of membrane associated proteins as bait, this method provides no information regarding the physical location of MORIP binding sites on the receptor. We therefore used the directed Y2H assay to first validate the interaction of the newly identified MORIPs with MOR, and then map the location and determine the specificity of the binding sites of these MORIPs on the receptor itself. Our studies showed that each of the MORIPs from the MYTH screen interacted specifically with the MOR-IL2 domain while several proteins from the traditional screen also interacted with additional intracellular portions of the receptor. Together, these mapping studies suggest that the MOR-IL2 represents a ''hot spot'' for protein/protein interactions. This is in accordance with our previous interaction studies using the D2 dopamine receptor (D2R), which showed that D2R-IL2 serves as a binding site for multiple interacting proteins [17,62]. Sequence comparisons indicate that the IL2 domains of the MOR,

KOR, and DOR are very highly conserved. Among the MORIPs tested, AUP1, Cx37, SIAH1, SIAH2, VAPA, and WLS also interacted with DOR and KOR in co-IP studies, suggesting that these MORIPs are likely to interact with the corresponding DOR and KOR-IL2 domains. As these protein interactions are not specific to any one subtype of opioid receptor it is likely that these proteins may be universal regulators of opioid function. Neither CSN5 nor DOK4 were found to interact with DOR or KOR. The failure of CSN5 and DOK4 to bind to DOR or KOR suggests that the non-conserved amino acids in the IL2 domains of these two opioid receptors may represent key residues that are necessary for the binding of CSN5 and DOK4 to the MOR. It is possible that these two MORIPs specifically regulate the function of MOR without affecting the function of DOR or KOR. The methods used to detect MOR/MORIP interaction in this paper were performed in the absence of agonist stimulation, which indicates that the interactions are ligand-independent. It is possible, however, that ligand stimulation would alter the MOR/MORIP interaction. This is the case with MOR and WLS where activation of the MOR with DAMGO or morphine enhances the MOR/WLS interaction [19]. The studies reported here add significantly to the list of known MOR binding partners, and as shown in Fig. 8, include proteins involved in virtually all aspects of MOR function including receptor-mediated signaling, trafficking, desensitization, cytoskeletal interaction, subcellular localization, degradation, and recycling. Of the novel proteins identified in our screens, we focused our attention on those of the DOK family of adapter proteins and those proteins involved in the ubiquitin/proteasome pathway. Two proteins from the downstream of tyrosine kinase (DOK) family, DOK4 and DOK5, were identified as MORIPs using a traditional Y2H screening approach. This family of proteins has a broad range of functions. In the nervous system, DOK4 has been shown to be required for axon myelination [63] and the regulation of GDNF-dependent neurite outgrowth (through activation of the Rap1-ERK1/2 pathway) [64]. DOK5 interacts with TrkB and TrkC neurotrophin receptors, and is involved in the activation of the MAPK pathway induced by neurotrophin stimulation [65]. Neurotrophins, like opioid receptors, have been linked to pain reception, reward, and synaptic plasticity [66,67]. For example, BDNF (a ligand of TrkB) has been implicated in substance dependence as well as antinociception [68,69,70,71]. It will clearly be of interest to determine whether there is a functional interaction between the MOR and neurotrophic factors in MAPK-mediated signaling, pain processing, reward, and/or memory formation. Several of the novel MORIPs identified in the current Y2H screens are components of the ubiquitin/proteasome pathway. Seven in absentia homologs 1 and 2 (SIAH 1 and 2) are RING finger E3 ubiquitin ligases that mediate ubiquitination and subsequent proteasomal degradation of target proteins [37]. CSN5 (COP9 subunit 5 or COP9S5) is a subunit of the large COP9 signalosome (CSN) which is responsible for the deneddylation and activation of Cullin RING E3 ubiquitin ligases that ubiquitinate target proteins for proteolysis [72]. Interaction of the MOR with these components of the ubiquitin/proteosome pathway was somewhat unexpected given the fact that the MOR is typically thought to be downregulated via lysosomal degradation [32]. However, recent studies have shown that upon internalization, the MOR is ubiquitinated in the first intracellular loop, and that this ubiquitination is involved in clathrin coated pit scission, endocytic sorting of the receptor, and down-regulation of receptor ligand binding [32,73]. Since SIAH1 and SIAH2 are known E3 ubiquitin ligases [74], and each is capable of interacting DADLE-induced decrease in MOR levels for each inhibitor treatment was compared to the percent reduction observed without inhibitor treatment using a two-sided paired Student's t-test. Error is expressed as standard error of the mean. * (p,.01) and ** (p,.005) indicate statistical difference as compared to no inhibitor treatment. IP indicates the antibody used for immunoprecipitation while IB indicates the antibody used for immunoblotting. doi:10.1371/journal.pone.0067608.g006 with the MOR, we asked whether either protein contributed to the ubiquitination or degradation of the MOR. Although our results suggest that both the lysosomal and the proteosomal pathways contribute to agonist-induced decreases in MOR expression levels, SIAH1 and SIAH2 do not appear play a role in this process. Overexpression of SIAH1, SIAH2, or truncated dominant negative forms of either protein, did not affect MOR protein or ubiquitination levels in unstimulated cells, agonist stimulated cells, or cells treated with lysosomal or proteasomal inhibitors. Recently, it has been shown that the E3 ligase, Smurf2 is involved in ligand- Figure 7. Role of SIAH1 and SIAH2 in regulating MOR ubiquitination. (A) Mapping the interaction site of MOR-IL2 on SIAH1 using GSTpulldowns. As controls, S-tagged constructs were incubated with untreated or GST-coated beads. (B) HEK-MOR cells were transfected with wild-type or truncated SIAH (trSIAH1 or trSIAH2) constructs and treated for 6 hours with 30 mM MG132. Proteins were immunoprecipitated and blots probed with either mouse anti-myc or mouse anti-HA to test for SIAH expression (left panels) and interaction with MOR (right panels), respectively. (C) HEK-MOR cells were transfected with wild-type or truncated SIAH constructs and either left untreated or treated with 10 mM DADLE for 6 hours. Blots were cut into sections and probed with rabbit anti-FLAG, mouse anti-myc, mouse anti-HA, or chicken anti-GAPDH antibodies. Bar graphs represent the average pixel density from 4 experiments normalized to GAPDH and untreated controls and subjected to a two-sided paired Student's t-test. None of the SIAH constructs caused significant changes (at p,.05) in steady-state levels of MOR protein expression or in DADLE-mediated decreases in MOR expression levels. induced ubiquitination of MOR [73]. Taken together, these studies argue against a role for the SIAH proteins in the regulation of MOR ubiquitination. Alternatively, it is possible that the MOR itself may serve as a scaffold to bring other MORIPs within the MOR signalplex into contact with the ubiquitination machinery. A potential example of this is the vesicle-associated protein synaptophysin, a previously identified MORIP that is a known ubquitination target of both SIAH1 and SIAH2 [51]. Knockout mouse experiments have shown that MOR is the primary mediator of the analgesic and rewarding properties of opioid drugs [1]. Since MORIPs regulate the functional output of the MOR, they represent potential candidates for proteins that may contribute to the underlying molecular mechanisms leading to drug addiction. Protein expression analysis of selected brain regions from morphine-treated mice indicate that the expression of DOK4, SIAH1, and WLS are altered in distinct brain regions following chronic morphine exposure. It is not known at this time whether these changes in expression are due to changes in protein stability, transcription, or translation. The MOR/WLS interaction, which was originally identified in the same MYTH screen reported here, has previously been functionally characterized in HEK-MOR cells by our laboratory [19,20]. We showed that morphine treatment promotes an increase in the MOR/WLS interaction at the plasma membrane. This in turn may be responsible for the decrease in Wnt protein secretion that occurs following morphine treatment. Thus it is tempting to speculate that the decrease in WLS expression observed in midbrain and striatum of mice after morphine administration could possibly lead to long term decreases in Wnt secretion within the CNS. Decreased Wnt secretion could therefore be responsible for many of the observed effects of chronic opioid exposure including decreased neurogenesis and decreased dendritic spine density that are known to be affected by Wnt signaling [75,76]. Although the functional relevance of changes in WLS, DOK4 and SIAH1 expression after morphine exposure is currently unknown, it is of interest that these changes in protein expression were observed within specific brain regions, rather than throughout the entire brain. Thus the changes in protein expression we detected may play an important role in the response to opioid drug exposure mediated by these brain regions. In the future, it should become possible to examine the role of these MORIPs in behavioral aspects of opioid addiction using animal models of opioid self-administration. The results presented in this study highlight the use of multiple screening methods to expand our knowledge of proteins that interact with, and contribute to the regulation of opioid receptormediated signaling in brain. Identification of the full panorama of MORIPs represents a critical step in understanding the normal regulation of the receptor life-cycle, as well as how receptor signaling or trafficking may be hijacked in response to drugs of [18,19,46,51,66,67,68,69,70]. MORIPs were grouped based on established functional properties. MORIPs identified in the current study are depicted by green shaded circles, while previously identified MORIPs are represented by white circles. Many of the newly identified MORIPs also interact with DOR and KOR and may be considered general ORIPS (see Results). doi:10.1371/journal.pone.0067608.g008 abuse. The MORIPs identified in this and other studies may also represent key players involved in other opioid-mediated processes such as neural development, nociception, aversion, and synaptic plasticity. As the list of MORIPs grows, and their contributions to receptor function become better understood, it is possible that some of these proteins will also become targets for new drug development to prevent and/or treat opioid addiction. A family of fluoride-specific ion channels with dual-topology architecture Fluoride ion, ubiquitous in soil, water, and marine environments, is a chronic threat to microorganisms. Many prokaryotes, archea, unicellular eukaryotes, and plants use a recently discovered family of F− exporter proteins to lower cytoplasmic F− levels to counteract the anion’s toxicity. We show here that these ‘Fluc’ proteins, purified and reconstituted in liposomes and planar phospholipid bilayers, form constitutively open anion channels with extreme selectivity for F− over Cl−. The active channel is a dimer of identical or homologous subunits arranged in antiparallel transmembrane orientation. This dual-topology assembly has not previously been seen in ion channels but is known in multidrug transporters of the SMR family, and is suggestive of an evolutionary antecedent of the inverted repeats found within the subunits of many membrane transport proteins. DOI: http://dx.doi.org/10.7554/eLife.01084.001 Introduction Fluoride pervades our biosphere, appearing in groundwater, sea, and soil typically at 10-100 μM levels (Weinstein and Davison, 2004). This ubiquitous inorganic xenobiotic inhibits two enzymes essential for glycolytic metabolism and nucleic acid synthesis: enolase and pyrophosphatase (Marquis et al., 2003;Samygina et al., 2007). Accordingly, many unicellular organisms, as well as green plants, use F − exporter proteins in their cell membranes to keep cytoplasmic F − concentration low, thereby minimizing the anion's toxic effects (Baker et al., 2012;Stockbridge et al., 2012). Two separate, phylogenetically unrelated families of F − exporters are now known: CLC F -type F − /H + antiporters, a subclass of the widespread CLC superfamily of anion-transport proteins, and a group of small membrane proteins known as the 'crcB' family or, as we rename them here (to avoid confusion with the CLCs), the 'Fluc' family. Fluc proteins are widespread among unicellular organisms. Deletion of the single Fluc gene in Escherichia coli produces hypersensitivity to F − , and this growth-phenotype may be rescued by F − exporter genes from a variety of bacterial species (Baker et al., 2012;Stockbridge et al., 2012). By expressing, purifying, and functionally reconstituting several bacterial Fluc homologs, we demonstrate that these are highly selective F − -conducting ion channels constructed as dimers of identical or homologous membrane-embedded domains arranged in an inverted-topology fashion. This type of molecular architecture is unprecedented among ion channels but is reminiscent of the dual-topology construction of small multidrug transporters (Rapp et al., 2006;Schuldiner, 2009;Morrison et al., 2012) and of the inverted structural repeats appearing in many membrane transport proteins (Forrest et al., 2010). in which La1 and La2 are joined together in tandem into single polypeptides. The purified proteins run cleanly on SDS-PAGE ( Figure 1-figure supplement 1) near the predicted molecular weights and migrate on size-exclusion columns (SEC) as monodisperse peaks about 0.6 ml ahead of the expected monomer position (Figure 1-figure supplement 2). Both Ec2 and Bpe are particularly stable, with chromatographic profiles unaltered even after several days in detergent solution at 37°C. Fluc is a four-TM membrane protein The predicted transmembrane topology of Fluc ( Figure 1B) was experimentally tested with Bpe, a naturally cysteine-free homologue poised for specific labeling at a unique cysteine substituted near the N-terminus (T3C). Lipid vesicles reconstituted with this mutant were treated with LysC protease to exclusively remove the C-terminal His tags exposed to the outside of the liposomes. Figure 1C shows that

Fluc proteins are randomly oriented in the liposomes. About half of the protein population has an externally exposed C-terminus and is susceptible to cleavage, as indicated by faster migration on SDS-PAGE. The other half is protected, with the C-termini exposed to the protease-inaccessible liposome interior. After proteolysis, a membrane-impermeant, thiol-reactive fluorophore was added to the liposome suspension to label externally exposed N-termini. This treatment labels only the lower band containing the polypeptides with externally exposed C-termini, thereby demonstrating that the N-and C-termini are exposed on the same side of the membrane, as anticipated by the Fluc hydropathy profile. To further confirm the predicted four-TM topology, similar experiments were performed with unique cysteines (N31C and E94C) placed in the first or third loops ( Figure 1C). Here, the label reacts exclusively with the proteins that retain the C-terminal His-tag, thus placing these loops and the C-terminus on opposite sides of the membrane. Fluc proteins form F − specific ion channels Purified Fluc proteins are functional, as seen in F − and Cl − efflux experiments in reconstituted liposomes , with ion-specific electrodes following the appearance of the anions in the external solution. Liposomes were reconstituted under 'Poisson-dilution' conditions, that is, with Ec2 or Bpe at protein densities so low that 30-50% of the liposomes are devoid of protein, and most of the protein-containing liposomes carry only a single functional unit (Walden et al., 2007). The liposomes eLife digest Fluorine is the thirteenth-most abundant element in the Earth's crust, and fluoride ions are found in both soil and water, where they accumulate through the weathering of rocks or from industrial pollution. However, high levels of fluoride ions can inhibit two processes essential to life: the production of energy by glycolysis and the synthesis of DNA and RNA bases. In polluted areas, organisms such as bacteria, algae and plants must remove fluoride ions from their cells in order to survive. Since ions cannot freely cross lipid membranes, organisms use proteins called channels or carriers to move ions into and out of their cells. Channel proteins form a pore, or channel, in the cell membrane, through which ions can quickly move from areas of high concentration to areas of low concentration. In contrast, carrier proteins can transport ions in both directions-that is, to and from areas of high concentration-but they are slower than channel proteins. A family of proteins that export fluoride from microbe and plant cells, thus allowing them to grow in the presence of this toxic ion, was discovered recently, but it was not clear if these proteins function as channels or as carrier proteins. Now, Stockbridge et al. find that these proteins, called Fluc proteins, are fluoride channels with an unusual architecture. Fluc proteins are found in many species of bacteria, and Stockbridge et al. show that a number of these, when purified and inserted into a lipid membrane, are channel proteins. Additionally, they do not transport related ions such as chloride, which means that they are unusually selective for ion channels. Two Fluc polypeptides associate to form a channel in the cell membrane, and Stockbridge et al. show that these two subunits are arranged in an antiparallel formation. Although this architecture is unprecedented among ion channels, it has been observed in carrier proteins in a range of organisms, and may indicate that Fluc proteins offer an evolutionary model for many carrier proteins. DOI: 10.7554/eLife.01084.002 were loaded with 150 mM each of KF and KCl and suspended in iso-osmotic solution containing 1 mM KF and KCl along with 300 mM K-isethionate (2-hydroxyethanesulfonate, a membrane-impermeant monovalent anion). Valinomycin (Vln), a K + ionophore, was then added to start the efflux by allowing the anions to move down their gradients. After protein-catalyzed ion efflux had reached completion, detergent was added to disrupt all the liposomes and reveal the protein-free fraction. Two striking results are apparent from the F − and Cl − efflux kinetics ( Figure 2A). First, the proteoliposomes are selectively permeable to F − ; Cl − efflux is undetectable on this timescale, and similar experiments in the absence of F − recapitulate this lack of Cl − transport. Second, Fluc-mediated F − efflux is so fast that an initial rate measurement is precluded by the 1-s dead-time of the stirred-cuvette system. An estimate based on a 35-nm liposome radius leads to a lower limit of ∼30,000 s −1 for the single-Fluc transport rate. Since this is substantially higher than turnover of any conformational-cycle based membrane transporter (Jayaram et al., 2008), it intimates that Fluc might be a F − channel. Accordingly, we recorded ionic currents mediated by Fluc-Ec2 inserted into planar phospholipid bilayers. Initial experiments aimed at maximizing the electrical signal used liposomes with high protein density, 10-50 Fluc copies per liposome. When fused into planar bilayers, these liposomes evoke a robust increase in bilayer conductance in discrete steps ( Figure 2B), each of which reports a single liposome inserting its packet of ion-conducting proteins. The heterogeneity in step-size reflects the wide distribution of liposome size (Miller and Racker, 1976;Accardi et al., 2004). After a few minutes, liposomes are perfused away to stabilize the bilayer conductance, thereby allowing examination of its properties. We applied a series of pulses to voltages from −100 mV to +150 mV and recorded the resulting currents to determine current-voltage (I-V) relations under various ionic conditions. As shown in the inset to Figure 2C, recorded in the presence of a salt gradient (300 mM NaF//30 mM NaF), the current appears as a 'leak' without any time-or voltage-dependent gating over a wide voltage range. The Fluc-mediated conductance is ideally selective for F − over Na + , as seen from the zero-current voltage (reversal potential) of 53 mV, indistinguishable from the Nernst potential for F − in the eightfold ionic activity gradient. High selectivity is further illustrated by the I-V relation under bi-ionic conditions (300 mM NaF//300 mM NaCl), whose high reversal potential (119 mV) sets a lower limit of ∼100 on the F − /Cl − permeability ratio. To detect single-channel currents, we prepared Fluc-reconstituted liposomes under Poisson-dilution conditions and fused them as above into planar bilayers. Figure 3A shows examples of Fluc-insertion events in three separate bilayers, along with a histogram of Fluc insertion-step amplitudes taken from ∼50 bilayers ( Figure 3B, upper panel). Most of these insertion-steps are 1.8 pA under these conditions, equivalent to 7 pS, while a minority (<20%) are 30-50% of this main step-size (illustrated in the third trace). The current shows occasional discrete fluctuations of two types: subsecond-timescale closing events and rare millisecond-timescale excursions to a substate of 50-60% of the main-state amplitude. The 'full-open' state persists >95% of the time at all voltages (−200 mV to +200 mV). Its high open probability further shows itself in an extended recording with three channels in the bilayer ( Figure 3A, red trace) and from the amplitude histogram of a typical single channel (right panel lower histogram). These channels display voltage-independent, anion-selective characteristics in harmony with the macroscopic currents reported above, and their appearance is strictly dependent on reconstituting Fluc protein into the liposomes. Accordingly, we conclude that these unitary current-steps reflect activity of single Fluc proteins, with single-molecule turnover of ∼10 7 F − ions/sec at −200 mV, a rate typical for ion channels and three to four orders of magnitude higher than the fastest known transporters. Thus, Fluc-Ec2 is a highly F − -selective channel, a protein with a transmembrane pore through which F − ion moves thermodynamically downhill by electrodiffusion. Fluc channels are dimers Fluc-Ec2 runs as a homodimer in detergent micelles on a size-exclusion column, as determined by combined measurements of UV absorbance, refractive index, and static light scattering of the eluting peak ( Figure 4A). Since an ion channel pore formed on a dimer axis is unprecedented, we also assessed Fluc's oligomeric state in its native habitat, the lipid bilayer. For this, we prepared Fluc-Bpe for singlemolecule total internal reflection fluorescence (TIRF) photobleaching experiments (Tombola et al., 2008) by labeling a unique cysteine mutant, R29C, with Cy5-maleimide and reconstituting the protein at a very low density to favor single-channel liposomes. Upon illumination of these liposomes immobilized on a glass surface, the covalently attached fluorophores bleach in several minutes ( Figure 4B,C), mostly in single and double steps ( Figure 4D). Given the measured labeling efficiency (72%), the preponderance of single-and double-bleach events is fully consistent with a dimer ( Figure 4E). Fewer than 5% bleach in three or four steps, consistent with co-localization of the liposomes, but not with trimeric or tetrameric architecture for the channel. The same conclusion was obtained in a second dataset using a different Bpe mutant, T3C (Figure 4-figure supplement 1). We further tested the oligomeric state of Fluc in membranes by exploiting the F − -transport function of the channel in 'Poisson-dump' experiments. The crux of this approach is the Poisson statistics of channel insertion into liposomes during reconstitution (Goldberg and Miller, 1991;Maduke et al., 1999;Walden et al., 2007); as an increasing amount of protein is reconstituted into a fixed amount of liposomes, the fraction of liposomes devoid of protein, f o , decreases exponentially. This protein-free fraction may be readily determined from F − efflux experiments, as above. Upon Vln addition, F − exits the liposomes that contain functionally active channels, but remains trapped within the protein-free fraction f 0 , which is determined by detergent addition ( Figure 4F): where ρ is the experimentally varied protein density (moles Fluc protein/mg lipid), M is the molecular weight of the Fluc subunit, n is the number of subunits in the active channel, and k is a constant dependent on the poorly known size and shape distribution of the liposomes. This constant is determined by calibrating the system with a channel of known molecular weight; for calibration we use Cl − efflux mediated by a high-turnover mutant of the Cl − /H + antiporter CLC-ec1-E148A/Y445A (Jayaram et al., 2008), a homodimer of 52 kDa subunits (Figure 4-figure supplement 2). Figure 4G shows f o curves expected for monomer, dimer, trimer, and tetramer architecture, along with experimental points from multiple preps, which coincide well with the dimer curve. The best exponential fit to the Fluc-Ec2 data corresponds to a molecular weight of 31.7 kDa for the functional channel, in unreasonably good agreement with the predicted size of the homodimer, 31.6 kDa. Similar results were obtained in parallel experiments with Fluc-Bpe (Figure 4-figure supplement 3). Fluc channels are built by dual-topology architecture In about half of the bacterial genomes in which it is found, Fluc appears as a single gene, as with the homodimeric channels Ec2 and Bpe ( Figure 5A). Other prokaryotes carry a pair of homologous Fluc genes arranged in tandem (∼25% sequence identity), strongly suggestive of primeval gene duplication. To test the function of paired Fluc proteins, we individually expressed, purified and reconstituted each of the twin Flucs of Lactobacillus acidophilus, La1 and La2 (Figure 1). Although the low yield of La2 precludes detailed biochemical analysis, functional examination of these fraternal twins produces an unambiguous result ( Figure 5B). Neither La1 nor La2 alone catalyzes F − efflux from liposomes, but co-reconstituting both proteins produces rapid F − efflux, indicating that the active channel requires heteromeric assembly. Comparison of Fluc sequences of singletons vs fraternal twins shows striking differences in charge distributions on the loops connecting the helices. The twin Flucs possess abundantly charged loops; for one of the pair, a positive charge bias resides on the N-and C-termini and loop 2, while its homologous partner exhibits a converse bias, with excess positive charge on loops 1 and 3 ( Figure 5D). Thus, according to the positive-inside rule for charge distribution in bacterial membrane proteins, (von Heijne, 1989) one of the twins is predicted to insert into the membrane with the termini in the cytoplasm, and the second should be oppositely oriented. In striking contrast, the singletons display relatively balanced charge distributions on either face. In a genomic analysis of small membrane proteins, including Fluc, Rapp and colleagues (Rapp et al., 2006) proposed that a balanced Arg+Lys distribution in loops indicates dual-topology oligomeric construction, in which subunits insert randomly in both transmembrane orientations, and those of opposite orientations associate together Figure 5C). Do singletons like Ec2 and Bpe find partners

in the membrane from a pool of randomly oriented Flucs to form inverted-topology homodimers? Do the fraternal Flucs find their homologous companions oppositely oriented in the membrane? A strong indication of antiparallel assembly appears in eukaryotic Fluc genomes. These genes all code for a pair of homologous Fluc sequences within a single open reading frame, and the linker connecting them contains a predicted TM helix ( Figure 5A). This 'inversion linker' would force the second Fluc domain to adopt a transmembrane topology upside down with respect to the first. These genomic characteristics of Fluc channels constitute compelling evidence for dual-topology construction, but since no previously known ion channels are built in this way, experimental support is needed to test such sequence-based suggestions. We approach the question in two ways. First, amino-group cross-linking of Fluc subunits is examined in detergent micelles. Even though the Ec2 subunit contains four amino groups, the dimer cannot be cross-linked by glutaraldehyde ( Figure 6A). Since these amines are all located on the same face of the protein-in loop 2 and both termini-glutaraldehyde would be expected to cross-link a conventional parallel dimer but not an antiparallel dimer. With an additional lysine substituted on the opposite face, in loop 1 (R25K) or loop 3 (N95K), glutaraldehyde treatment now leads to the appearance of a dimer band on the PAGE gel ( Figure 6A). Similarly, lysine-less Bpe is not cross-linked by glutaraldehyde, despite its solvent-accessible N-terminal amino group. Single lysine mutations on the side opposite to the N-terminus, in loop 1 (R29K) or loop 3 (R95K), beget a cross-linked dimer ( Figure 6B), whereas solvent-accessible lysines on loops 2 (R68K), on the C-terminus (R130K), or on both do not ( Figure 6B). These results are in natural harmony with antiparallel orientation and are difficult to understand in terms of a parallel-topology dimer, but do not by themselves provide rigorous proof. A second test of the dual-topology idea applies the fused-dimer strategy of Schuldiner and colleagues (Steiner-Mordoch et al., 2008;Nasie et al., 2010) to the Fluc La1/La2 heterodimer. We expressed several constructs in which La1 and La2 are fused together with linkers designed to force the two domains into parallel or antiparallel orientations ( Figure 6C). For the antiparallel construct, Laf-TM, the domains are linked with a non-dimerizing glycophorin-A TM helix (Lemmon et al., 1992;Fleming et al., 1997). Guided by eukaryotic Fluc sequences, we made loop 4 long and hydrophilic (18 residues), and loop 5 short (8 residues). The domains of the parallel constructs, LapA and LapB, are connected by hydrophilic linkers of 14 or 26 residues, respectively. All three fused constructs were expressed, purified, and reconstituted into liposomes, Laf-TM giving the best yield and chromatographic monodispersity (Figure 1-figure supplement 2). Both parallel constructs have minimal transport activity, as shown for LapA ( Figure 6D). In marked contrast, Laf-TM behaves quantitatively in Poisson-dump experiments like a fully active monomeric F − channel ( Figure 6E,F) with high selectivity for F − over Cl − . This result strongly implies that functional Fluc channels are built with the two 4-TM subunits or domains in antiparallel transmembrane topology, an arrangement that prohibits axial symmetry along the pore ( Figure 6C). Discussion In this work, we have taken four initial steps towards understanding the molecular character of the Fluc family of F − exporters. The results demonstrate that the mechanism by which Fluc transports F − is thermodynamically passive electrodiffusion through a transmembrane channel, that the channel is unusually F − -selective, and that it is constructed as a dimer of subunits or homologous domains. In addition, the two subunits in the functional dimer are very likely oriented antiparallel to each other within the membrane. How could a passive channel move F − out of the cell against an environmental F − challenge? It might seem that an active transport mechanism would be required, as in bacteria that use CLC F -type F − /H + antiporters. However, Fluc channels can work to effectively expel F − in two ways. First, the negative membrane potential maintained by many classes of metabolizing cells tilts the anion's equilibrium towards expulsion. Second, the high pK a of HF (∼3.4), anomalous among haloacids, dictates that in weakly acidic environments, a significant amount of extracellular HF is present (Baker et al., 2012). This membrane-permeant acid readily enters the cell and dissociates at the higher cytoplasmic pH. F − would therefore accumulate far above its extracellular concentration if no conductive pathway for the anion were present in the membrane, but a passive F − channel would undermine this weak-acid accumulation effect. A deeper enigma is Fluc's remarkable discrimination between F − , which is so difficult to desolvate, and Cl − , which is orders of magnitude more abundant in the environment. Liposome flux experiments establish a far more dramatic F − /Cl − selectivity than the 100-fold lower limit set by electrical recording. Anion-efflux assays measure Cl − turnover <30 ions/s, while in the same assays with F − , unitary turnover rates exceed 30,000 s −1 , and direct single-channel recording yields a turnover of 10 6 -10 7 F − /sec. These values conservatively put F − /Cl − selectivity >10 4 -fold, the highest value known between close analogs for any ion channel. It is this formidable selectivity that allows a bacterium or yeast to express a constitutively open F − channel in its energy-coupling membrane while avoiding a massive Cl − conductance that would otherwise catastrophically collapse its membrane potential. High ion selectivity requires that F − be largely desolvated as it passes through the Fluc channel (Hille, 2001), an energetically demanding step (>100 kcal/mol) that has been an obstacle to the development of F − -selective small-molecule ligands (Cametti and Rissanen, 2009). This chemical problem motivates future work to identify selectivity regions in Fluc sequences and, ultimately, structures. The molecular architecture inferred here is unconventional and unexpected. Most ion channel proteins are built on a barrel-stave plan in which three to seven similar subunits surround an axial pore running perpendicular to the membrane plane, the wider pores requiring a larger number of subunits. But an ion-conduction pore formed by a two-stave barrel-a dimer-is unprecedented until now. We are obliged to point out, however, that Fluc's high open probability precludes a definitive conclusion that the observed single-channel currents are mediated by a single pore rather than two in parallel, especially in light of the infrequent subconductance ( Figure 3). However, this substate is not precisely half the condunctance of the fully open channel, and the protein's small size makes a two-pore dimer, along the lines of CLC channels (Middleton et al., 1994;Ludewig et al., 1996;Middleton et al., 1996), unlikely, but a picture like this is not firmly ruled out at our investigation's early stage. A final surprise that provides an intriguing glimpse into the evolution of membrane transport proteins is the dual-topology construction inferred for Fluc. Unprecedented among known ion channels, dual topology has been observed in the SMR family of proton-coupled multidrug antiporters such as EmrE (Schuldiner, 2009(Schuldiner, , 2012Morrison et al., 2012), and it recalls the inverted structural repeats that provide the mechanistic scaffolds of many transporters (Forrest et al., 2010). Oppositely oriented repeats within subunits are rare in ion channels, and so it is unclear whether this architecture provides particular mechanistic advantages for Flucs. The inverted structural repeats of transporters almost certainly evolved from gene duplication and subsequent fusion of sequences coding for dimeric proteins with an odd number of TM helices (Zuckerkandl and Pauling, 1965;Forrest and Rudnick, 2009). The Fluc family obligingly provides examples in modern genomes of all stages along such an evolutionary pathway, including duplicated gene-pairs in bacteria and fused homologues in eukaryotes. Fusion of the latter genes must have required evolutionary gymnastics to capture a TM helix as an inversion linker for maintaining antiparallel topology of subunits with an even number of membrane crossings. Finally, we note a conspicuous analogy to Fluc construction: the aquaporin channel. This protein's narrow, water-permeable pore is formed within a single 6-TM subunit on the interface between two structurally similar domains in inverted orientation with respect to each other (Harries et al., 2004). It is thus entirely feasible to construct a narrow pore at a dual-topology dimer interface. The Fluc-aquaporin analogy is also of interest since the desolvated F − ion, with a radius of 1.3 Å, is almost the same size as H 2 O. Design of Fluc constructs Constructs used in this study are summarized in Table 1. Synthetic gene constructs for Fluc-Ec2, Fluc-Bpe, Fluc-La1, and Fluc-La2 were inserted into a pASK vector with a C-terminal LysC recognition site and hexahistidine tag (TRKAASLVPRGSGGHHHHHH) (Maduke et al., 1999). The fused constructs Laf-TM, LapA, and LapB were formed from the sequence of La1 followed by a linker sequence (Table 1) leading into the sequence of La2 (without the first methionine), to which a C-terminal hexahistidine tag was appended. Laf-TM contained a transmembrane linker composed of a non-dimerizing glycophorin A helix (Lemmon et al., 1992). Site-directed mutagenesis was performed using standard PCR techniques. All mutants were confirmed functionally active in a liposome-based F − efflux assay. Topology determination 10 mM stocks of AlexaFluor 647-C2-maleimide and fluorescein succinimidyl ester were prepared in anhydrous DMSO and kept in the dark at −80°C until use. The Bpe construct contained unique cysteine mutations T3C, N31C, or E94C for maleimide labeling, with an R29K background to increase labeling by the succinimidyl ester. Proteoliposomes (POPC/POPG, 20 mg/ml) were prepared with 2 μg Bpe/mg lipid, with an intraliposomal solution of 300 mM NaF, 25 mM HEPES pH 7.5, and 1 mM cysteine. A 100-μl sample of the liposomes was treated with LysC (0.05 U, 1 hr, 22°C) to completely cleave all externally exposed C-terminal His tags and was then centrifuged through a 1.5-ml G-50 column equilibrated with 300 mM NaF, 25 mM HEPES pH 7.5 to remove cysteine and LysC from the external solution. AlexaFluor maleimide (50 μM) was then added to label external cysteine residues, and after an hour freshly prepared phenylmethylsulfonyl fluoride (PMSF) was added to quench residual LysC, and 10 mM cysteine to quench the maleimide. Liposomes were disrupted by 150 mM β-octylglucoside, and fluorescein succinimidyl ester (50 μM) was added for 1 hr to stain protein amino groups before quenching with 10 mM Tris. Samples were run on 10-20% SDS-PAGE gels and fluorescent bands were visualized using a Typhoon 9410 Variable Mode Imager (GE Healthcare). Cysteine-conjugated protein was visualized with the red laser (633 nm excitation, 670 nm emission), and fluorescein-conjugated total protein with the blue laser (488 nm excitation, 526 nm emission). Fluorescein was used for staining total protein because of the large amount of lipid in the sample, which interferes with Coomassie or silver stain. Anion transport assays Anion efflux assays were performed as described previously . In short, liposomes prepared with 10 mg/ml E. coli polar lipids containing 150 mM KF, 150 mM KCl, and 25 mM HEPES pH 7 were extruded 21 times through a 400-nm membrane filter and passed over a 1.5-ml G-50 Sephadex column equilibrated in external buffer composed of 300 K-isethionate, 25 mM HEPES pH 7, 1 mM KF or KCl, according to the ion measured. Liposomes were diluted 20-fold into external buffer in a stirred cuvette, and efflux was initiated by addition of 1 μM K + ionophore valinomycin. After ∼30 s, 50 mM β-octylglucoside was added to disrupt the liposomes. Cl − or F − appearance in the external solution was monitored using homemade Ag/AgCl or Cole-Parmer LaF 3 /EuF 3 electrodes, respectively; under the ionic conditions used here, these electrodes are ideally selective for Cl − or F − , and they show no cross-reactivity towards the other anion. In some experiments, F − efflux was monitored by 90° light scattering at 600 nm in a fluorimeter. A liposome sample containing 300 mM KF, 25 mM HEPES-KOH pH 7 was diluted 200-fold into 2 ml of a degassed isotonic solution containing 300 mM K-or Na-isethionate, 25 mM HEPES-KOH pH 7 in a stirred cuvette, and efflux was initiated by 1 μM Vln. Water efflux maintaining osmotic balance leads to time-dependent shrinking and flattening of the initially spherical liposomes, accompanied by a ∼10% increase in 90° light scattering (Jin et al., 1999;Stockbridge et al.,

2012). After transport was complete, p-trifluoromethoxyphenyl hydrazine (FCCP) was added to the cuvette, effectively making the protein-free liposomes specifically permeable to F − to provide a quantitative measurement of the fraction of liposomes, f 0 , free of Fluc channels. For Poisson-dump experiments, data were collected for proteins from four independent preparations of CLC-ec1 and Fluc-Ec2, and two independent preparations of Fluc-Bpe, Fluc-Laf, Fluc-LapA, and Fluc-LapB. Planar lipid bilayer recording Planar bilayer recording was as previously described (Accardi et al., 2004), using a POPE/POPG phospholipid mixture to form the bilayer and 150 mM NaF-1.5% agar salt bridges to connect the recording chambers to the Ag/AgCl electrodes through 1 M KCl wells. Liposomes (0.5 μl added to the cis solution) were fused into the bilayer, with 300 mM or 30 mM NaF, 15 mM Mops-NaOH pH 7, in the cis or trans chamber respectively, trans defined as electrical ground. Ionic conditions for recording were established by perfusion of either chamber. Macroscopic I-V relations were determined with families of voltage pulses (0.5-1 s) from a holding potential of zero to command-voltages from −100 to +150 mV in 10-mV increments. Current output from a Axopatch 200A amplifier was low-pass filtered at 500 Hz and sampled at 2 kHz in pCLAMP software. Voltages were corrected for junction potential, <3 mV. Signals for single-channel analysis were subsequently filtered digitally at 100 Hz. SEC-SLS-UV-RI The molecular mass of Fluc in detergent was determined using the static light scattering/refractive index method (Folta-Stogniew, 2006;Slotboom et al., 2008) implemented at the W. M. Keck Foundation Biotechnology Resource Laboratory, Yale University. Briefly, purified protein was passed over a Superdex-75 SEC column equilibrated in 400 mM NaF, 10 mM MES pH 6.5, 5 mM DM and coupled in-line with light scattering, refractive index, and UV detectors. Ovalbumin, transferase, bovine serum albumin, and carbonic anhydrase were used as standards. The molecular mass was determined in two ways: by the three-detector method (Hayashi et al., 1989), and from light scattering intensity at several angles (ASTRA software, Wyatt Technology Corp.). Single-molecule TIRF photobleaching For the labeling reaction, freshly prepared Bpe with a single cysteine substitution (T3C or R29C) was incubated with a 20-fold molar excess of Cy5-maleimide for 1 hr at room temperature in the dark and quenched with a 100-fold molar excess of cysteine. To wash away excess fluorophore, the protein was bound to cobalt affinity beads and washed with 45 column volumes of cobalt wash buffer. The protein was eluted with 400 mM imidazole and the labeling efficiency was calculated from the protein absorbance at 280 nm (ε Bpe = 38,960 M −1 cm −1 , calculated from protein sequence), and the Cy5 absorbance at 655 nm (ε Cy5 = 2.5 × 10 5 M −1 cm −1 ), with spillover correction of 0.017. Parallel experiments with wildtype, cysteine-free Bpe showed that the background labeling was <2% under the same conditions. Protein was reconstituted into liposomes at very low density (0.15 µg/mg lipid). Flow chambers were prepared using two glass coverslips that were cleaned by sonication in 2% micro-90 detergent (Cole-Parmer, Vernon Hills, IL), ethanol and 0.1 M KOH. Liposomes were formed immediately prior to analysis. BPe T3C liposomes were sonicated vigorously for 10 min until clarity, yielding ∼30 nm liposomes. BPe R29C liposomes were extruded 21 times through a 100 nm filter, then 21 times through 30 nm filter (Avestin, Ottawa, Canada) to form small vesicles. Liposomes were diluted ∼10 5 -fold in reconstitution buffer, and loaded into the flow chamber, where they spontaneously adhere to the glass surface. The fluorescence from each liposome was measured using a custom throughobjective TIRF microscope built for single molecule detection (Friedman et al., 2006). Cy5 fluorescence was measured by illuminating the sample with a helium-neon 633 nm laser (75 µW power) and collecting the emitted light through a 633 nm long-pass filter (1 s per frame). Each spot was centered inside a 3 × 3 pixel area using linear interpolation and drift-correction, and the intensity was integrated over the duration of the experiment. Primary Fallopian Tube Carcinoma Primary fallopian tube carcinoma (PFTC) is a rare tumor. It is seen in the age group between 40 and 60 years. It has resemblance to epithelial ovarian cancer, both clinically and histologically. Only 5% of patients show symptoms of profuse watery vaginal discharge (hydrops tubae profluens). Ultrasound, computerized tomography, and magnetic resonance imaging (MRI) can help in arriving at a diagnosis. Tumor marker cancer antigen (CA)-125 levels can be used as a diagnostic and prognostic marker to detect recurrence. The most common histopathology is serous papillary. Transcoelomic exfoliation of cells is the most common mode of spread of the tumor. Other modes of spread include contiguous invasion, hematogenous and transluminal dissemination. The PFTC is managed surgically like epithelial ovarian cancer. Adjuvant chemotherapy is also similar to epithelial ovarian cancer. INTRODUCTION Primary fallopian tube carcinoma has an incidence of 0.14 to 1.8% of malignancy of female genital tract. 1 The exact etiology of this tumor is not known, but hormonal, reproductive, and genetic factors increase the risk. High parity and oral contraceptive use decrease the risk. The disease has no relation to age, parity, race, infertility, endometriosis, and smoking. 2 It is most commonly seen between 4th and 6th decade of life. However, there are case reports of the disease in young girls aged 17 to 19 years. 3 Profuse and intermittent serosanguinous vaginal discharge, colicky pain relieved on discharge, and abdominal or pelvic mass (Latzko's triad) are seen in 15% of patients. Profuse serosanguinous discharge (hydrops tubae profluens) may be seen in 3.5% of patients. Transabdominal and transvaginal ultrasound can help in making a diagnosis. Color Doppler may help in detection of neovascularization in the adnexal region. Irregularity of tubal walls, such as papillary projections, pseudo septae, and vascular abnormalities like microaneurysms, arteriovenous shunts, tumor lakes, and dichotomous branching are seen in malignant tumor on three-dimensional scan. 4 The MRI helps in detection of tumor infiltration into bladder, vagina, pelvic side walls, pelvic fat, and rectum. More than 80% of patients have an elevated CA-125 levels. The CA-125 levels in the preoperative period is an independent risk factor for disease-free survival and overall survival of patients with PFTC. Response to chemotherapy can also be assessed by CA-125 levels. 5 CASE REPORT Forty-eight-year-old Mrs X was admitted in our institution, Government Medical College, Kozhikode, Kerala, India, with abdominal pain and distention. Patient had no history of anorexia, weight loss, menstrual abnormalities, or discharge per vaginum. She was Para 3 with 2 living children with history of postpartum sterilization. She was known hypertensive on medication. On examination, she was obese with BMI of 35 with stable vitals. Her abdominal examination revealed an abdominopelvic mass about 16 weeks of gravid uterus. On examination, the cervix was hypertrophied, uterus was enlarged, and there was a mass felt through right fornix higher up. Ultrasound of abdomen and pelvis revealed uterus enlarged, measuring 14 × 07 × 12 cm, myometrium heterogeneous with multiple fibroids, largest on right lateral wall 83 × 81 mm. Left ovary was normal and right ovary not visualized separately. There was a mixed echogenic mass with solid and cystic component 12 × 5 × 5 cm with septations, suggestive of complex adnexal mass. Magnetic resonance imaging confirmed multiple fibroids and a large tubular mass lesion superior to uterus suggestive of right tubular mass. The CA-125 was elevated to 175.1 U/mL. Staging laparotomy was done on Feb 3, 2016 under regional anesthesia. Findings: No ascites, uterus enlarged to 12 weeks size with multiple fibroids, right tube enlarged appearing retort-shaped measuring 20 × 6 cm, smooth surface with intact serosa and no extension to adjacent structures, Right ovary, left tube, and ovary were normal (Fig. 1). Normal saline was instilled and fluid taken for cytology. Surgery performed was total abdominal hysterectomy with bilateral salpingo-oophorectomy with infracolic omentectomy and pelvic lymphadenectomy. Postoperative, patient recovered well. Histopathology report: No malignant cells in the cytology. Right fallopian tube enlarged measuring 20 × 8.5 cm with tumor occupying area of 6 × 5 cm, extending to wall of fallopian tube, serosa free. Findings of serous adenocarcinoma high grade. Bilateral ovaries and left tube were histologically normal. Uterus showed leiomyoma with proliferative endometrium. Omentum was histologically unremarkable. Lymph nodes showed reactive hyperplasia. Hence, the tumor was staged as stage IA (tumor nodes metastases Classification T1a). Since it was high grade, she was advised adjuvant chemotherapy (Carboplatin AUC6 plus Paclitaxel 175 mg/m 2 ) every 3 weeks for 3 to 6 cycles. The tumor can be staged as shown in Table 1. Surgery is the treatment of choice in stages I to IV fallopian tube cancers. A very small percentage of women with epithelial fallopian tube cancers can be treated with surgery alone. Most patients with stage I disease need adjuvant chemotherapy. Surgical staging and tumor debulking are the goals in management. The National Comprehensive Cancer Network guidelines suggest administering 3 to 6 cycles of chemotherapy for stages IA to IC disease, and 6 to 8 cycles for stages II to IV disease. 6 A response rate of 70% is seen in patients with adjuvant platinum-based chemotherapy with 12.5 months of median response duration. Postoperative radiation is to be avoided, as it has low efficacy and more complications. Paclitaxel is also emerging as an important adjuvant in advanced fallopian tube carcinomas. Adjuvant chemotherapy is not needed in patients who are in early-stage disease (stages IA and IB). Two randomized controlled trials, i.e., ICON 1 and ACTION, compared mere observation vs platinum-based chemotherapy. They reported a 5-year survival rate of 74% without chemotherapy vs 82% with platinum-based chemotherapy. 7 CONCLUSION Primary fallopian tube carcinoma is a rare tumor and accounts for <1% of all female genital cancers. It is a disease of nulliparous women and serous papillary histology is more often seen. Surgery is the main modality of management and it should comprise of total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and pelvic and para-aortic lymphadenectomy. Advanced disease needs extensive debulking. The tumor responds to platinum-based chemotherapy. Prognostic factors include stage and residual tumor. Patients with stage I (low-risk patients) undergoing optimal surgical cytoreduction may not need adjuvant chemotherapy. However, those patients with low-risk stage I disease with inadequate surgical staging and those with high-risk stage I or II disease should receive 3 to 6 cycles of adjuvant chemotherapy with carboplatin plus paclitaxel. Treatment of persistent/recurrent disease needs second-line treatment based on platinum-free interval. Localized, late relapse may need secondary cytoreduction in a selected few patients. Prenatal and perinatal analgesic exposure and autism: an ecological link Background Autism and Autism Spectrum Disorder (ASD) are complex neurodevelopmental disorders. Susceptibility is believed to be the interaction of genetic heritability and environmental factors. The synchronous rises in autism/ASD prevalence and paracetamol (acetaminophen) use, as well as biologic plausibility have led to the hypothesis that paracetamol exposure may increase autism/ASD risk. Methods To explore the relationship of antenatal paracetamol exposure to ASD, population weighted average autism prevalence rates and paracetamol usage rates were compared. To explore the relationship of early neonatal paracetamol exposure to autism/ASD, population weighted average male autism prevalence rates for all available countries and U.S. states were compared to male circumcision rates – a procedure for which paracetamol has been widely prescribed since the mid-1990s. Prevalence studies were extracted from the U.S. Centers for Disease Control and Prevention Summary of Autism/ASD Prevalence Studies database. Maternal paracetamol usage and circumcision rates were identified by searches on Pub Med. Results Using all available country-level data (n = 8) for the period 1984 to 2005, prenatal use of paracetamol was correlated with autism/ASD prevalence (r = 0.80). For studies including boys born after 1995, there was a strong correlation between country-level (n = 9) autism/ASD prevalence in males and a country’s circumcision rate (r = 0.98). A very similar pattern was seen among U.S. states and when comparing the 3 main racial/ethnic groups in the U.S. The country-level correlation between autism/ASD prevalence in males and paracetamol was considerably weaker before 1995 when the drug became widely used during circumcision. Conclusions This ecological analysis identified country-level correlations between indicators of prenatal and perinatal paracetamol exposure and autism/ASD. State level correlation was also identified for the indicator of perinatal paracetamol exposure and autism/ASD. Like all ecological analyses, these data cannot provide strong evidence

of causality. However, biologic plausibility is provided by a growing body of experimental and clinical evidence linking paracetamol metabolism to pathways shown to be important in autism and related developmental abnormalities. Taken together, these ecological findings and mechanistic evidence suggest the need for formal study of the role of paracetamol in autism. Background Autism Spectrum Disorder (ASD) is a severe developmental disorder defined by social deficits, communication deficits, repetitive behaviors and fixated interests that appear in early childhood [1,2]. Despite a large and rapidly expanding body of literature, the etiology of ASD remains poorly understood. There is substantial evidence implicating oxidative stress, inflammation and immune dysregulation, although no single coherent explanation has emerged [3]. Recent twin studies provide evidence that susceptibility to ASD may have significant environmental components, in addition to genetic heritability [4,5]. Several lines of evidence suggest that medication use in pregnancy and early childhood may play a role in ASD etiology. Specifically, Torres and, Becker and Schultz, have hypothesized that paracetamol (acetaminophen, N-acetylp-aminophenol, or APAP) has increased ASD risk [6,7]. It has been shown that autistic children have a decreased capacity to sulfate paracetamol, which is the primary metabolic route for children [8]. When paracetamol is metabolized through the alternative routes it has been shown in humans to induce oxidative stress and immune dysregulation [9]. A recent investigation found transcriptomic changes in full-genome human miRNA expression indicating, for the first time, immune modulating effects and oxidative stress responses to paracetamol even at low doses [10]. Studies in animals have shown paracetamol to induce apoptosis and neurotoxicity [11,12]. Several studies hypothesize increased apoptosis in the autistic brain [13][14][15][16]. Paracetamol is one of the most common antipyretic analgesic medicines worldwide. In 1980, after sufficient evidence emerged of an association between salicylates and Reyes syndrome, paracetamol essentially replaced aspirin as the primary treatment of fever in children and pregnant women [17,18]. Since that date, paracetamol consumption throughout the world has increased dramatically [19]. Paracetamol sales in the United States (US) have had a continual upward trend. In 1982, US paracetamol sales were approximately $400 million; by 2008 they had risen to $2.6 billion [20,21]. Although prevalence data for autism and ASD are of uncertain accuracy, many authors report strong increases in prevalence over this same time period. Theoharides and colleagues for example reported prevalence prior to 1980 as approximately 4/10,000 and Baio et al. estimated that US autism/ASD prevalence to have risen to about 110/10,000 today [22,23]. Observing the correlation between two parallel time trends is of limited inferential utility; however the paracetamol -ASD link is strengthened by an observation first made by Becker and Schultz [7]. In 1982 and again in 1986, product tampering led to a few bottles of a leading brand of paracetamol tablets being contaminated with cyanide. In each case, a rapid and brief decline in paracetamol sales occurred, with the long term upward trend recovering within a year. In three populations for which good data are available, Becker and Schultz noted that brief dips in the rising autism prevalence curves mirrored these sales anomalies. The prevalence curves continued their upward trend after 1988. Study aims Several lines of evidence suggest that the etiologically relevant period for the development of ASD may be in utero or possibly in early infancy [24,25]. We sought additional evidence to test the hypothesis that use of paracetamol in pregnancy or in early childhood might be a risk factor for ASD. Studies have reported population variation in prescribing patterns and usage rates which has allowed us an opportunity to investigate the correlation between prenatal exposure to paracetamol and autism spectrum disorder prevalence. Additionally, we noted that analgesic prescribing habits for neonates and infants changed in the mid 1990's with a shift in perspective on neonatal pain which has afforded us an additional opportunity to look at population variation in analgesic use [26]. Research beginning in the 1980's documented the negative consequences associated with inadequate treatment of pain in children [27][28][29]. Pain guidelines specifically for children were established in 1992 by the Agency for Health Care Policy and Research, in 1998 by the World Health Organization and in 2001 by the American Academy of Pediatrics [30][31][32][33]. A common neonatal medical procedure is circumcision, which typically occurs during the postpartum hospital stay, within the first two days of life for a vaginal delivery and first four days for a cesarean section [34]. Prior to the 1990's circumcision was generally performed without analgesics. A 1994 study by Howard et al. found that when paracetamol is given regularly every 6 hours for at least the first 24-hour postoperative period, infants demonstrated decreased responses to pain [35]. This study lead to the development of circumcision pain management guidelines by the American Academy of Pediatrics [36] and others [37][38][39]. These guidelines include the suggestion of a first dose of paracetamol two hours prior to the procedure, and doses every 4-6 hours for 24 hours following the procedure. Thus newborn males often receive 5-7 doses of paracetamol during the developmentally vulnerable initial days of life. Variations in circumcision frequency in different populations allowed us an additional approach to investigate the correlation between paracetamol use and ASD prevalence. This hypothesis seemed particularly relevant in light of the approximately 4.6 times higher prevalence of autism in males compared to females [23]. Methods Two separate analyses were conducted. The first examined the association between prenatal paracetamol exposure and ASD prevalence using maternal usage data as a proxy for prenatal exposure. The second analysis investigated the association between circumcision rates as a proxy for early life paracetamol exposure in males and ASD prevalence. Data for investigating prenatal paracetamol exposure and autism prevalence To investigate the relationship of ASD to prenatal exposure to paracetamol, population maternal paracetamol usage rates were compared to population autism prevalence rates for as many countries as available data permitted. The population autism prevalence rates utilized were from studies reported in the U.S. Centers for Disease Control and Prevention Autism Prevalence Summary Table [40]. The maternal paracetamol usage rates by country were drawn from studies identified by a systematic search of the peer reviewed medical literature using the National Library of Medicine's Pub Med database. Autism prevalence rates The Autism Prevalence Summary Table from the Center for Disease Control website [40] summarized the results of 59 prevalence studies conducted worldwide. This table recorded the author, year published, country, time period studied, age range studied, number of children in the population, the identification criteria, the methodology used and the prevalence. The oldest and youngest birth years in each study were calculated based on the time period and age range studied. Each study was extracted to verify the table results and to identify the ratio of males to females to calculate the male prevalence of autism/ASD. Generally, the CDC Prevalence Summary Table reported the more narrowly defined autism rate rather than the more inclusive diagnosis of autism spectrum disorder (ASD). The CDC Prevalence Summary reported the ASD rate for the Kim 2011 study in South Korea; however for consistency with all other results, the more narrowly defined autism rate was extracted and utilized in this analysis [41]. Two older studies referred to in the CDC table could not be located and were excluded (Lotter et al. 1966) (Brask et al. 1970). Two additional study were excluded, the first because of incomplete case ascertainment [42] and the second because of lack of total population data [43]. A total of 55 studies from the CDC summary were utilized in this analysis (Additional file 1). Prenatal paracetamol usage rates Maternal paracetamol usage rates were extracted from studies examining the use of paracetamol in pregnancy and relationships to other outcomes. A search of the English language literature on Pub Med was conducted for the past 20 years with a search date of April 18, 2012 using the terms prenatal, maternal, pregnancy, acetaminophen, paracetamol, medication, drugs, analgesic, pain relief, over the counter, and the different country names. Various combinations of the terms were used to maximize the results. If a study appeared to be relevant it was extracted and reviewed to identify a maternal paracetamol usage rate. If data were found only for overall analgesics, nervous system drugs, all over the counter medications or all western pharmaceuticals an assumption was made that paracetamol use represented 80%, 80%, 70% and 60%, respectively. These proportions were established a priori, conservatively approximated based on the findings of the US National Birth Defects Prevention Study and used for all studies regardless of country [18]. This literature search yielded 33 studies with medication usage rates for 14 out of 17 countries with autism prevalence rates. Two studies were excluded because they were subsets of two included studies to yield a total of 31 studies [44,45]. If multiple studies were identified for a country a summary usage rate was calculated using the weighted average by study population size. The characteristics of the prenatal medication usage studies are summarized in Additional file 2. Because of concerns of changing autism prevalence rate over time, an a priori decision was made to restrict the analysis to include autism prevalence studies in which the range of birth years had at least one year of overlap with the range of birth years of the prenatal paracetamol usage studies. If multiple autism prevalence studies met this criterion for a given country, a weighted average based on study population size was calculated. This reduced the number of prenatal exposure studies used in this analysis to 20. There were inadequate data available to conduct a U.S. state level analysis. Data for investigating early life exposure to paracetamol for circumcision and autism prevalence To investigate the relationship of early life paracetamol exposure for male neonates to autism spectrum disorder, population circumcision rates were compared to male population autism prevalence rates for two time periods. Male autism prevalence rates calculated from studies reported by the Center for Disease Control in the Summary of Autism/ASD Prevalence Studies (described above) were compared to male circumcision rates from studies identified by a systematic search of the peer reviewed medical literature using the National Library of Medicine's Pub Med database. An additional U.S. state level analysis was done with available data (limited to the more recent time period) by comparing state and time period stratified male autism prevalence rates from the U.S. studies from the CDC Autism Prevalence Summary Table to newborn circumcision rates from the Health Care Utilization Project [46]. Circumcision rates The circumcision rates were obtained by systematic search of the peer reviewed medical literature using the National Library of Medicine's Pub Med database. A search of the English language literature on Pub Med was conducted using the terms circumcision and the different country names. The circumcision rates utilized the best identified information. If infant circumcision rates were available, they were utilized over national rates. If changing rates were presented, the rates for years closest to the study birth years were utilized. When a published paper was not available the rate was estimated. The estimation was calculated using the same methodology as the World Health Organization and the Circumcision Independent Reference and Commentary Service, calculated from the sum of the numbers of Jewish and Muslim males [47,48]. Data for the percentage of Jews by country were obtained from the Jewish Virtual Library [49]. Data for the percentage of Muslims by country were obtained from a Pew Forum report [50]. While most Jewish and Muslim males are circumcised, true circumcision rates are unknown and circumcision rates based upon religion are only an approximation. Both the World Health Organization and the Circumcision Independent Reference and Commentary Service indicated that this would likely underestimate the true prevalence (see Additional file 3). Annual U.S. state level infant circumcision rates were available for eight years between 1997 and 2006 from the Health Care Utilization Project (HCUP) of the Agency for Health Care Research and Quality (AHRQ) of the United States Department of Health and Human Services. Thirteen states had complete data for this eight year period and an additional seventeen states had partial data [46,51]. For the data analysis, the studies of autism prevalence were divided into two time periods. The first consisted of all prevalence studies in which all subjects were born before 1995 (35 country level studies). The assumption is that during this time period

paracetamol would not generally have been administered during the circumcision procedure. The second, post-1995 cohort includes prevalence studies in which some subjects were born after 1994 (1995+), (20 country level studies) when circumcision pain was first recognized and treated. The assumption is that during this time period some portion of the cohort would have been administered paracetamol during the circumcision procedure. For countries with multiple studies in a time period, a summary prevalence was calculated using a weighted average based on study population size (Additional file 3). An additional U.S. state level analysis was conducted for the post-1995 cohort. Similarly, for states with multiple studies in a time period, a summary prevalence was calculated using a weighted average based on study population size. State level circumcision data were not available to conduct a pre-1995 analysis [Additional file 4]. Each of the data sets was checked for normality using standard graphical and statistical methods. Within the limits of these small datasets, the normality assumption was not seriously violated and so Pearson's parametric correlation coefficient was used with an information weighted (1/variance) linear regression model. Prenatal exposure For the country-level analyses, synchronous data were available from 8 countries. A country's average prenatal paracetamol consumption was found to be correlated with its autism/ASD prevalence (r = 0.80, Figure 1 and Additional file 5). The trend among the 8 countries indicates that a change of 10% in population prenatal paracetamol usage was associated with an increased autism population prevalence of 0.53/1000 persons (95% CI: 0.13 to 0.93) (Figure 1). Early life exposure A strong correlation (r = 0.98) was found in the countrylevel data between circumcision and autism spectrum disorder prevalence rates for boys born after 1995 (when circumcision guidelines began recommending analgesics). The slope of this trend for the 9 countries with available data indicates that a change of 10% in the population circumcision rate was associated with an increase in autism/ ASD prevalence of 2.01/1000 persons (95% CI: 1.68 to 2.34) (Figure 2). Data were available for 12 countries for boys born before 1995 (Figure 3), and the trend in the data was weaker; the correlation between circumcision prevalence and autism/ ASD prevalence was still good (r = 0.89), but the slope of the trend was only a sixth of that for the later period for a 10% change in circumcision rate, there was an increase in autism/ASD prevalence of 0.35/1000 persons (95% CI: 0.22 to 0.47). Across all country-level studies prior to the widespread use of paracetamol for circumcision (all born prior to 1995), the weighted average autism/ASD male to female prevalence ratio was 3.9 to 1. For the post 1995 cohort, this ratio increased to 5.6 to 1. Available data allowed a parallel analysis of U.S. states post-1995, but not for the earlier period. The data for the 14 U.S. states with available data show a remarkably similar pattern to the 8 countries (this set of countries does not include the U.S. to avoid double counting) for the same time period (Figure 2, r = 0.95). Discussion These ecological analyses identified positive correlations between autism/ASD prevalence and indicators of both prenatal and very early life paracetamol exposures. If these patterns are confirmed in formal epidemiologic studies, the use of paracetamol during pregnancy and at the time of circumcision may help to explain autism/ ASD prevalence variations between the sexes, among countries, and within U.S. states and ethnic groups. The Figure 2 Graph of country and U.S. state-level data on Autism/ASD and circumcision prevalence rates. All studies in which at least some cohort members were born after 1995. This graph includes country-level studies with the U.S. stratified to state-level studies for the post-1995 cohort (no overall U.S. data point). The assumption is that, due to changes in neonate prescribing practices in the mid-1990s, some cohort members in each study would likely be exposed to paracetamol at the time of circumcision. Autism rates are population weighted averages of all studies for a country or U.S. state. (See Additional files 1, 3, 4 and 5). Figure 3 Graph of country-level data on Autism/ASD and circumcision prevalence rates. All studies in which all cohort members were born before 1995. The assumption is that all of the cohort members in each study would not likely be exposed to paracetamol at the time of circumcision. U.S. state level data was not available for this pre-1995 cohort. Autism rates are population weighted averages of all studies for a country. (See Additional files 1 and 3). close to six-fold difference in rate of change in autism/ ASD prevalence before and after the recognition of pain at the time of circumcision is suggestive of a possible effect caused by the shift to the use of paracetamol. The change in the overall time period average male to female autism/ASD prevalence ratio (weighted by study size) from 3.9 to 1 in the unexposed time period to 5.6 to 1 in the second time period with a probable paracetamol exposure may also be suggestive an effect from this exposure. Limitations It is important to acknowledge that this analysis has numerous and significant limitations. First and foremost, correlation is not causation and as such no causal inference is intended. Homogeneity of exposure and prevalence assessment methodologies among the studies has been assumed, but each may be subject to misclassification, confounding and bias. The change in autism/ASD prevalence, circumcision prevalence and paracetamol usage rates over time may not have been adequately addressed. Circumcision rates are presented as a proxy for an early male neonatal exposure to paracetamol. However, this assumption is not without significant limitations. The timeline for the actual implementation of child pain management protocols and the utilization of paracetamol with circumcision is not known. Additionally, pain management guidelines suggest that paracetamol alone is not sufficient to manage circumcision pain so a nerve block or local anesthesia may also be administered, which may be confounding factors. In general, this type of ecologic study has significant limitations that severely limit causal inference. Ecologic bias or the failure of ecological associations to correspond to biologic effects at the individual level is a concern. It has been shown that the relations seen in country level data may poorly reflect the relationships that exist on an individual basis [52]. Despite these limitations, the consistent patterns reported here support the need to further investigate this potentially important hypothesis. Prenatal exposure trends Previous research has identified paracetamol usage trends that curiously coincide with the rise in prevalence and population demographics of autism/ASD. In the US Slone Epidemiology Center Birth Defects study paracetamol was the most commonly used medication amongst all subjects with usage higher during pregnancy than before pregnancy. In the early 1980's about 42% of women used paracetamol during the first trimester of pregnancy. The rate climbed to over 65% in the early 1990's, where it has essentially remained through 2004 [18]. Maternal viral infection requiring hospitalization during the first trimester and maternal bacterial infection in the second trimester have been associated with diagnosis of ASD in the offspring (Hazard ratios 2.98 (95% CI: 1.29 to7.15) and 1.42 (95% CI: 1.08 to1.87), respectively) [53]. In a recent study, maternal self-reported influenza was associated with a twofold increased risk of infantile autism and a febrile episode lasting more than seven days was associated with a threefold increased risk [54]. Each of these maternal infections or a prolonged febrile episode would likely increase the exposure to paracetamol. In the U.S., usage of paracetamol by pregnant women mirrors the population demographics of women whose children develop autism spectrum disorder, by race, age and education [18,55,56]. The population demographics for mothers who circumcise their children are also very similar, with rates increasing with socio-economic status and insurance coverage rates [57][58][59]. Studies have shown that a parent's own usage rates of paracetamol and other medications correlate with what they give to their children, so a similar demographic usage pattern would be expected for childhood exposure [60,61]. This synchronous U.S. pattern may be suggestive of an additive nature of both prenatal and early life exposure to paracetamol and a relationship to autism/ASD (Figure 4). Early life exposure trends Paracetamol is the most common drug administered to US children and the predominant analgesic/antipyretic drug among children up to 24 months of age [62]. Paracetamol is suggested for pain management following vaccinations. In 1983 the average U.S. child received 8 immunizations before age 2. In 2011, the average was 25, a 313% increase [63,64]. From the perspective of the current hypothesis, these represent increased opportunities for paracetamol exposure in pain management (although administering several vaccines at once means analgesia may not increase proportionally). A recent study representing one-fifth of all pediatric hospital admissions in the U.S., identified paracetamol as the most common drug administered to children over one year of age and the second most common drug administered for those under one year; more than 40% of hospital stays in both groups include paracetamol [65]. Hospitalization of children and neonates for infection as well as non-infectious disease have been associated with increased risk of ASD in a large Danish cohort (hazard ratios 1.38 (95% CI: 1.31 to 1.43) and 1.76 (95% CI: 1.68 to 1.86), respectively) [66]. Biologic plausibility Paracetamol has four important metabolic pathways ( Figure 5). The two main pathways are glucuronidation and sulfation. Paracetamol is mainly metabolized in the liver via conjugation with glucuronide and sulfate and then excreted. Both these metabolic routes yield inactive, non-toxic final products. Glucuronidation is the primary metabolic pathway in adults and sulfation is the primary pathway for paracetamol metabolism until age 10-12 years [67]. Neonates, in general, have lower capacity to metabolize drugs due to the underdevelopment of the glucuronidation pathway and inefficiency and immaturity in renal function [68]. Three studies of neonates with postnatal age ranging from 1-3 days obtained a glucuronidation/sulfation (G/S) ratio between 0.12 and 0.28 [69][70][71]. This is in contrast to 11 month old children with a significantly higher G/S ratio of about 0.7 and adults with a G/S ratio approaching 2.0 [71,72]. Low birth weight and bilirubinemia have also been found to reduce glucuronidation capacity, both of which have been associated with autism [73][74][75][76]. Autistic children have been shown to have abnormal sulfate capacity and have been shown to have a specific inability to sulfate paracetamol [8,77,78]. Parents of autistic children have also been shown to have abnormal transsulfuration metabolism [79]. When the capacity to metabolize through the primary pathways is depleted or saturated, the fraction of the dose converted to reactive metabolites increases and the secondary metabolic pathways become increasingly involved [80]. One of the two secondary pathways is cytochrome P-450 (CYP P-450) mediated, forming a highly reactive metabolite, n-acetyl-p-benzoquinoneimine (NAPQI) which reacts with cellular glutathione (GSH) to form a non-toxic conjugate, which is subsequently excreted. Once GSH is exhausted, NAPQI binds to cellular proteins, including mitochondrial proteins reducing the ability to detoxify, which can lead to oxidative stress, immune system activation, hepatocellular death, nephropathy and asthma [10,19,81,82]. It has been shown that paracetamol treatment induces greater glutathione depletion in male mice [11,83]. Alterations in GSH homeostasis have been a consistent observation among autistic children and their mothers [84][85][86][87]. Also, decreased glutathione levels have been associated with preeclampsia [88]. In the Danish Birth Cohort, women who used paracetamol during the third trimester of pregnancy had increased risk of preeclampsia RR = 1.40 (95% CI: 1.24 to 1.58) [89]. In two studies, maternal preeclampsia has been associated with increased risk of having a child with ASD (OR = 1.69 (p = .0005) and RR = 1.64 (95% CI: 1.08 to 2.49), respectively) [90,91]. Additionally, during pregnancy a women's sulfation capacity is reduced which may precipitate activation of immune responses, via this P-450 pathway [92,93]. The activation of immune response and pro-inflammatory cytokine interleukin signaling has recently been detected even at therapeutic doses of paracetamol [10,94]. Converging evidence highlights the important role of many of the same cytokines in mediating maternal immune activation effects on the neurodevelopment of autistic offspring [95][96][97][98][99][100][101][102][103][104][105]. A fourth metabolic pathway, accounting for about 6% of paracetamol metabolism, has been identified that is believed to be related to the mechanism of analgesic action [106]. This pathway involves deacetylation of paracetamol in the liver producing p-aminophenol that conjugates with arachidonic acid in the brain and in the spinal

cord [107,108]. P-aminophenol is recognized to be involved in paracetamol nephrotoxicity [81,109]. More recently, P-aminophenol has been shown to produce a significant loss in mouse cortical neuron viability at therapeutic concentrations [12]. This suggests another possible pathway for a neurotoxic effect of paracetamol when the principle metabolic routes are exhausted. In summary, there are several lines of evidence that suggest that prenatal and or early life paracetamol exposure may adversely affect neurodevelopment. Prenatal exposure may trigger maternal immune activation with possible effects on fetal brain development. In early life, maturational compromises to the glucuronidation pathway at the time of the circumcision related exposure, in combination with the compromises to the sulfation pathway that typify autistic children, may lead to utilization of the suboptimal secondary metabolic routes with the potential for adverse neurological effects in susceptible individuals [110][111][112]. Conclusions In this hypothesis generating exploratory analysis, several lines of evidence support the plausibility of a relationship between prenatal and early life exposure to paracetamol and autism spectrum disorder. It is proposed that the use of paracetamol in pregnancy and/or early childhood may alter immune processes increasing the risk of autism spectrum disorder in susceptible individuals. In an ecologic analysis, with all the previously discussed limitations, a correlation was found between maternal prenatal use of paracetamol and autism spectrum disorder. Additionally, a correlation was identified for the first time between neonatal circumcision with a probable paracetamol exposure and autism spectrum disorder. These relationships along with the synchronous rise in use of paracetamol and ASD, the convergence of the potential biologic mechanisms and the identification of plausible causes of increased male susceptibility provide consistent evidence of an association. Large scale population based epidemiologic studies are needed to confirm or disprove this association. Additional files Additional file 1: All Studies Reported in CDC Summary of Autism/ ASD Prevalence Studies. Summary of CDC studies divided into two groups. The first group has some members born post-1995 (some likely to be exposed to paracetamol at time of circumcision) and the second group born pre-1995 (exposure unlikely). The chart displays the study author, country, year published, range of birth years, ratio of males to females, study population, and the overall and male ASD prevalence . Additional file 2: Prenatal Paracetamol Exposure Studies Summary. All prenatal paracetamol studies identified by a systematic search of the PubMed database. Studies without overlapping birth years with the autism prevalence studies are denoted with an "*" and were not utilized in the analysis. The chart displays the study author, country, year published, range of birth years, the study population, paracetamol usage rate, and study methodology . Additional file 3: Summary of Weighted average Autism studies and Circumcision Rates by Country. Country weighted averages of CDC autism studies from Additional file 1. Data is divided into the two cohorts, pre and post −1995. The chart displays the circumcision rates and sources for each country and the study weights. The data in this table is used for the country-level analysis and comparison between the pre and post −1995 cohorts [193][194][195][196][197][198][199][200][201][202][203][204][205][206]. Additional file 4: U.S. Studies Stratified by State from CDC Summary of Autism/ASD Prevalence Studies. The six U.S. autism prevalence studies from Additional file 1 in the post 1995 cohort stratified by state. Chart displays the study author, country, year published, range of birth years, ratio of males to females, study population, the circumcision rates and sources, and the overall and male ASD prevalence. * Analysis restricted to 2006 data of 8 year olds. Additional file 5: Summary of Weighted average Autism Prevalence and Prenatal Exposure to Paracetamol Data. The Autism Prevalence Studies from Additional file 1 that had overlapping cohort birth years with the prenatal paracetamol studies in Additional file 2 are summarized. The study author, country, year published, range of birth years, population, overall ASD rate, the summary weighted average ASD rate by country, the weighted average APAP exposure rate,the APAP study range of birth years and the study weights. This data was used for Figure 1 and to calculate the prenatal correlation and trend. Competing interests The authors declare that there are no conflicts of interest. Authors' contributions AB conceived of the study, made substantial contributions to its design and coordination, acquired the data, performed the data analysis and drafted the manuscript. DK made substantial contributions to concept and design, and revised the manuscript critically for intellectual content and data interpretation. Both authors read and approved the final manuscript. Life-Threatening Severe Hyperkalemia Presenting Electrocardiographic Changes Hyperkalemia is a common and potential life-threatening electrolyte disorder in patients presenting to the emergency setting. It is known that renal insufficiency may abate the toxic effects of hyperkalemia on electrocardiographic abnormality formation. Herein, I have reported a case of severe hyperkalemia associated with renal insufficiency in a woman patient that presented to the emergency department with weakness and hypotension. Initial electrocardiogram showed accelerated AV junctional rhythm. Incipient treatment for hyperkalemia was begun and afterwards hemodialysis was performed for patient. Then, electrocardiogram was repeated that primary changes were removed and vital sign was stable. Eventually, it is better that hemodialysis starts from commencement for patients with life-threatening severe hyperkalemia associated with acute or chronic renal failure. Introduction Hyperkalemia is a common and potential life-threatening electrolyte disorder in patients presenting to the hospital setting. The prevalence of hyperkalemia varies in diverse studies, with estimates ranging between 1% to 10% of hospitalized patients [1]. Mild hyperkalemia (serum potassium concentrations of 5.5-6.5 mmol/L) has been associated with tall peaked T waves; moderate hyperkalemia, with loss of P waves (6.5-7.5 mmol/L) and QRS complex widening (7.5-8.0 mmol/L); severe hyperkalemia (8.0-10.0 mmol/L), with ventricular arrhythmias and asystole [2]. Other electrocardiographic manifestation has been reported in the literature [3]. It can cause lethal arrhythmias, flaccid paralysis, and respiratory difficulty. When changes detected on electrocardiogram, severe hyperkalemia should be treated urgently, even before laboratory confirmation, especially in critically ill patients [4,5]. It is known that chronic renal failure may detract the toxic effects of hyperkalemia on electrocardiographic abnormality formation [6,7]. In this study has been reported a case of life-threatening severe hyperkalemia with electrocardiogram changes that rapidly resolved this changes with treatment. Case Report A 76 year old woman patient presented to the emergency room with progressive generalized weakness since 12 days ago. Patient was anuric from 2 days ago. Associated symptoms includes anorexia, nausea, vomiting and confusion. She had a history of chronic renal insufficiency and hypertension for 3 years ago and also chronic heart failure from 3 months ago. There was no history of dialysis during this period. Also, drug history was negative. Patient was ill in appearance. In the emergency room, initial blood pressure was 80/60 mmhg and pulse rate was 70 per min. Other examinations were normal. The electrocardiogram showed accelerated AV junctional rhythm (Figure 1). Metabolic acidosis (pH of 7.12 and base excess of −11.6 mmol/L) was noted through arterial blood gas analysis. Routine laboratory studies reveal severe hyperkalemia (serum potassium, 8.4 mmol/L) and renal dysfunction (urea, 342 mg/dl and serum creatinine, 9.5 mg/dL). Serum levels of cardiac enzymes were within normal limit. She was promptly treated with intravenous fluid, 20 mL of intravenous 10% calcium gluconate, 10 units of intravenous insulin with 50% dextrose and 84 meq/l sodium bicarbonate intravenously. Then, temporary hemodialysis was performed for 3 h through a right internal jugular vein instantly before preparing results of laboratory tests. After 2 h of management, a repeat electrocardiography was normal without primary changes (Figure 2), his vital sign was stable with blood pressure 110/65 mmHg, and serum potassium level was 4.3 mmol/L. Metabolic acidosis was partially improved with pH of 7.29 and base excess of −3.5 mmol/L. Her renal function was progressively improved with serum creatinine, 2.3 mg/dL. Discussion Hyperkalemia is a life-threatening electrolyte abnormality resulting in a perilous cardiac arrhythmia. The electrocardiographic manifestations of hyperkalemia include the peaked T waves generally considered the earliest sign, progressive prolongation of the PR and QRS intervals and decreased amplitude and eventual loss of the P waves. As serum potassium level rises, sinoatrial and atrioventricular conduction was blocked, causing escape rhythms. This is followed by widening of the QRS complex and merging with the T wave to form a "sine-wave" appearance, which may result in ventricular fibrillation or asystole [8]. In this case, typical electrocardiographic features of severe hyperkalemia (>8.0 mmol/L) including the absence of P waves and junctional escape rhythm was noted. Prompt treatment of hyperkalemia was imperative to prevent further development of fatal cardiac arrhythmia. First, intravenous calcium gluconate should be infused in patients with electrocardiographic changes to stabilize membrane potential [9]. Then insulin with dextrose infusion, intravenous or nebulized beta-adrenergic agonists (salbutamol), or intravenous sodium bicarbonate can be used to increase potassium shift from extra-to intracellular space [10]. In this case, immediately, performed empirical treatments of hyperkalemia based on his electrocardiographic manifestations before achieving results of serum electrolyte concentration. Prompt inception with intravenous calcium, insulin with glucose, sodium bicarbonate and short-term hemodialysis completely improved hyperkalemia and recovered junctional rhythm with hemodynamic instability within the first few hours. It is also important to identify and eliminate underlying causes or precipitating factors of hyperkalemia. The common risk factors of hyperkalemia include acute or chronic renal insufficiency, medications interfering with urinary potassium excretion, and excessive intake of potassium containing diets [10]. Hyperkalemia can result in life-threatening cardiac arrhythmias and patients with underlying renal impairment would be predisposed to its development. Prompt and aggressive management with medical therapy and hemodialysis could be life-saving in patients with severe hyperkalemia Presented with marked arrhythmia and hemodynamic instability. Figure 1 The electrocardiographic changes on admission of patient that showed the absence of P waves and junctional rhythm. Figure 2 A repeat electrocardiograph was performed after 2 h of management for hyperkalemia that was normal and primary changes was removed. Characterization of the Interaction between Protein 4.1R and ZO-2 Multiple isoforms of the red cell protein 4.1R are expressed in nonerythroid cells, including novel 135-kDa isoforms. Using a yeast two-hybrid system, immunocolocalization, immunoprecipitation, and in vitro binding studies, we found that two 4.1R isoforms of 135 and 150 kDa specifically interact with the protein ZO-2 (zonulaoccludens-2). 4.1R is colocalized with ZO-2 and occludin at Madin-Darby canine kidney (MDCK) cell tight junctions. Both isoforms of 4.1R coprecipitated with proteins that organize tight junctions such as ZO-2, ZO-1, and occludin. Western blot analysis also revealed the presence of actin and α-spectrin in these immunoprecipitates. Association of 4.1R isoforms with these tight junction and cytoskeletal proteins was found to be specific for the tight junction and was not seen in nonconfluent MDCK cells. The amino acid residues that sustain the interaction between 4.1R and ZO-2 reside within the amino acids encoded by exons 19–21 of 4.1R and residues 1054–1118 of ZO-2. Exogenously expressed 4.1R containing the spectrin/actin- and ZO-2-binding domains was recruited to tight junctions in confluent MDCK cells. Taken together, our results suggest that 4.1R might play an important role in organization and function of the tight junction by establishing a link between the tight junction and the actin cytoskeleton. Multiple isoforms of the red cell protein 4.1R are expressed in nonerythroid cells, including novel 135-kDa isoforms. Using a yeast two-hybrid system, immunocolocalization, immunoprecipitation, and in vitro binding studies, we found that two 4.1R isoforms of 135 and 150 kDa specifically interact with the protein ZO-2 (zonula occludens-2). 4.1R is colocalized with ZO-2 and occludin at Madin-Darby canine kidney (MDCK) cell tight junctions. Both isoforms of 4.1R coprecipitated with proteins that organize tight junctions such as ZO-2, ZO-1, and occludin. Western blot analysis also revealed the presence of actin and ␣-spectrin in these immunoprecipitates. Association of 4.1R isoforms with these tight junction and cytoskeletal proteins was found to be specific for the tight junction and was not seen in nonconfluent MDCK cells. The amino acid residues that sustain the interaction between 4.1R and ZO-2 reside within the amino acids encoded by exons 19 -21 of 4.1R and residues 1054 -1118 of ZO-2. Exogenously expressed 4.1R containing the spectrin/actin-and ZO-2-binding domains was recruited to tight junctions in confluent MDCK cells. Taken together, our results suggest that 4.1R might play an important role in organization and function of the tight junction by establishing a link between the tight junction and the actin cytoskeleton. Erythrocyte protein 4.1R is an 80-kDa cytoskeletal protein critical in circulating red cells for

the dynamic organization and maintenance of the spectrin/actin cytoskeleton and for the attachment of the cytoskeleton to the cell membrane through interactions with integral membrane proteins such as glycophorin C and band 3 (1). In nonerythroid cells, an additional and prevalent 135-kDa 4.1R isoform class has been detected. It contains a 209-amino acid extension at its N terminus end (2). Multiple isoforms of 4.1R exist. They arise through complex alternative pre-mRNA splicing pathways (2,3), post-translational modifications (4), and use of at least two translation initiation sites (5,6). In nucleated erythroid and nonerythroid cells, a spectrum of proteins ranging from 30 to 210 kDa has been detected that contains epitopes for 4.1R (7,8). Unlike the strict peripheral distribution of 4.1R in mature erythrocytes, 4.1R isoforms in nonerythroid cells are localized in the nucleus and nuclear matrix (9 -13), at points of cell-cell or cell-matrix contact (14,15), and in centrosomal and Golgi structures (16,17). In addition, 4.1R may also interact with microtubules and stress fibers (14,18). However, the precise identity of 4.1R isoforms in these subcellular structures, their binding partners, and their biological significance in nonerythroid cells are not well understood. Like 4.1R, many members of the protein 4.1 superfamily, including ezrin, radixin, moesin, merlin, and myosin, are known to associate with the plasma membrane (reviewed in Ref. 31). A Drosophila homolog of protein 4.1, Coracle, is localized at septate junctions in ectodermally derived epithelial cells (32). Drosophila Neurexin and DLG (Discs Large), homologs of proteins that are known to bind to 4.1, glycophorin C, and human DLG, respectively, are also localized at the septate junctions (33,34). Collectively, these studies suggest the localization of 4.1 to the septate junction, an invertebrate specific junction with molecular components analogous to the vertebrate tight junction (TJ) (35). In addition to membrane binding activity, members of this family participate in important cell signaling events. For example, merlin is involved in growth regulation (36), and ezrin, radixin, and moesin are involved in Rho-dependent signaling pathways (37,38). To explore the function of 4.1R isoform(s) in nonerythroid cells, we searched for their binding partners. In this study, using a combination of molecular/genetic, biochemical, and immunofluorescence studies, we establish that two nonerythroid 4.1R isoforms bind to the C-terminal proline-rich domain of X-104, a human homolog of the canine protein ZO-2 (henceforth referred to as hZO-2), through its C-terminal domain. ZO-2 belongs to the membrane-associated guanylate kinase (MAGUK) family of proteins and is known to organize the TJ in association with other TJ-associated proteins (39 -41). We report here that these isoforms of 4.1R colocalize with ZO-2 and occludin and also form an intracellular complex with ZO-2, ZO-1, occludin, actin, and ␣-spectrin at MDCK cell TJs. Our results suggest that in nonerythroid cells, 4.1R may have a role in the organization and function of the TJ by establishing a direct link between the TJ and the actin cytoskeleton. For domain mapping, plasmids carrying inserts fused to Gal4-BD or Gal4-AD were cotransformed into Y190 and assayed for ␤-galactosidase activity on nitrocellulose filters. In some cases, liquid ␤-galactosidase assays using o-nitrophenyl ␤-D-galactopyranoside or chlorophenol red-␤-D-galactopyranoside were used as described in the CLONTECH manual. Cell Culture and Transient Transfection-MDCK cells (American Type Culture Collection CRL 1772) were used in this study because they are known to express TJ proteins and to form TJs at confluent density (47). We have described the culture conditions for MDCK cells and have documented the expression of 4.1R in this cell line at both the protein and RNA levels (13). Transfection was performed using Lipo-fectAMINE reagent (Life Technologies, Inc.) according to the manufacturer's protocol. Cells were plated on poly-D-lysine-coated coverslips in a six-well tissue culture plate 1 day prior to transfection. Cells at 50 -70% confluency were transfected with 2 g of plasmid DNA. After 36 -48 h, cells on coverslips were removed, fixed, and processed for immunofluorescence staining with anti-ZO-1 Ab as described below. Rabbit preimmune serum equivalent to 50 g of IgG was added to 2 mg of proteins and incubated for 2 h at 4°C, followed by the addition of 200 l of a 50% suspension of protein A-Sepharose CL-4B (Amersham Pharmacia Biotech) in immunoprecipitation buffer for 2 h at 4°C. The supernatant was collected by centrifugation at 3000 ϫ g for 5 min at 4°C and split equally into seven tubes. Preimmune serum, affinitypurified anti-HP Ab, anti-ZO-1 Ab, anti-ZO-2 Ab, anti-occludin Ab, rabbit IgG, or anti-p53 Ab (an irrelevant antibody as a control) containing 6 g of IgG was added to different tubes and incubated for 2 h at 4°C. Protein A-Sepharose CL-4B (100 l of a 50% suspension) was then added and incubated overnight at 4°C on a rocking platform. Following incubation with protein A-Sepharose, the samples were centrifuged at 3000 ϫ g for 10 s at 4°C and washed 10 times with 1 ml of immunoprecipitation buffer. The samples were resuspended in 80 l of SDS sample buffer (62.5 mM Tris-HCl, 10% glycerol, 2% (w/v) SDS, 5% 2-mercaptoethanol, and 10 g/ml bromphenol blue, pH 6.8) and boiled in a water bath for 5 min. The samples were centrifuged at 10,000 ϫ g for 15 min at 4°C, and the supernatants were fractionated on 6 -12% SDS-polyacrylamide gels. Transfer of proteins to nitrocellulose or polyvinylidene difluoride membranes and detection of 4.1R, ZO-1, ZO-2, occludin, actin, and spectrin were carried out by immunoblotting using an ECL detection kit (Amersham Pharmacia Biotech) as described earlier (13). Quantitation of proteins from chemiluminograms was done with NIH Image software for the Apple Macintosh computer. To visualize proteins, the gels were stained with GelCode ® SilverSNAP TM from Pierce. In Vitro Binding Assay-Recombinant 4.1R/GST proteins were expressed and affinity-purified by coupling to glutathione-Sepharose beads as described (13). The recombinant proteins were quantitated by Coomassie Brilliant Blue staining using bovine serum albumin as a standard. Desired [ 35 S]methionine-labeled segments of hZO-2, hZO-1, and human occludin were in vitro translated using a TNT ® SP6 Quick Coupled transcription/translation system (reticulocyte lysate system; Promega) according to the manufacturer's recommendations. The translation products were quantitated based on percent of methionine incorporation and 35 S-specific activity. For in vitro binding, equal amounts of labeled proteins were incubated with 20 g of 4.1R/GST fusion proteins coupled to glutathione-Sepharose beads for 2 h at 4°C in 300 l of binding buffer (50 mM Tris, pH 7.4, 50 mM NaCl, 10 mM sodium pyrophosphate, 1.5 mM sodium orthovanadate, 4 g/ml aprotinin, 4 g/ml leupeptin, 4 g/ml antipain, 12.5 g/ml chymostatin, 12 g/ml pepstatin, 130 g/ml ⑀-aminocaproic acid, 200 g/ml paminobenzamidine, and 1 mM phenylmethylsulfonyl fluoride). After incubation, the beads were washed 10 times with 1 ml of binding buffer containing 0.5% Triton X-100 each time. The beads were then resuspended in 30 l of SDS sample buffer, boiled for 5 min, and analyzed by SDS-polyacrylamide gel electrophoresis (12% gel; National Diagnostics, Inc., Atlanta). The gels were treated with Enlightning TM (NEN Life Sciences Products), and binding of 35 S-labeled proteins was detected by fluorography. To determine the dissociation constants between 4.1R and ZO-2, ZO-1, or occludin, 5 g of 24-kDa/GST or GST (as a control) coupled to glutathione-Sepharose beads was incubated with 0.0025-0.015 g of 35 S-labeled translated products as described above. After extensive washing, the bound proteins were eluted with 50 mM Tris-HCl containing 20 mM glutathione, pH 8.0. The amount of bound and free labeled protein was determined from 35 S-specific activity and adjusted for binding to the GST-conjugated Sepharose beads. A Scatchard plot of the data was generated. Each experiment was repeated three times for estimation of K d . Fluorescence and Confocal Microscopy-MDCK cells were grown on poly-D-lysine-coated coverslips to confluence. The coverslips were washed three times with PBS and fixed in freshly prepared 2% paraformaldehyde in PBS for 25 min at room temperature. Cells were washed with PBS and permeabilized with a solution of 0.2% Triton X-100 and 10% normal goat serum in PBS for 15 min. The cells were blocked with 10% normal goat serum and 50 mM NH 4 Cl for 1.5 h at room temperature and washed three times with PBS containing 0.3% bovine serum albumin. Cells were incubated with primary antibodies (1:100 dilution of anti-occludin Ab, 1:100 dilution of anti-ZO-2 Ab, 1:200 dilution of anti-ZO-1 Ab, 1:50 dilution of anti-10-kDa Ab, 1:50 dilution of anti-16-kDa Ab, and 1:25 dilution of anti-HP Ab) for 1 h at room temperature and washed three times as described above. They were then incubated with a 1:100 dilution of Texas Red-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories, Inc.) or a 1:50 dilution of fluorescein isothiocyanate-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories, Inc.) at room temperature for 1 h and washed three times again as described above. When double staining was desired, after the first set of staining, cells were blocked with unconjugated goat anti-rabbit Fab fragment (10 g/ml in PBS; Jackson ImmunoResearch Laboratories, Inc.) for 1 h at room temperature and washed three times as described above. Incubation with the second set of antibodies was the same as with the first set of antibodies. Mounting of the coverslips and processing of images using a Noran confocal laser scanning image system and Intervision software (Noran Instruments Inc., Middleton, WI) were the same as described previously (13). Immunoelectron Microscopy-Confluent MDCK cells were washed with PBS and fixed in 4% paraformaldehyde and 0.1% glutaraldehyde in PBS overnight at 4°C. The cells were then infiltrated with 15% polyvinylpyrrolidone and 1.95 M sucrose in PBS (48), and ultrathin cryosections were cut at Ϫ100°C with a Reichert FCS cryoultramicrotome (Leica Inc.). Cryosections were stretched using 2.3 M sucrose and mounted on Formvar/carbon-coated grids. Sections were blocked with 10% normal goat serum and 20 mM glycine in PBS for 30 min. Sections were sequentially incubated for 60 min with anti-ZO-2 (1:20 dilution) or anti-HP (1:5 dilution) antibody and AuroProbe EM goat anti-rabbit IgG (1:50 dilution; Amersham Pharmacia Biotech). The sections were postfixed in 2% glutaraldehyde in PBS. The immunogold-labeled sections were examined in an electron microscope (EM 10, Carl Zeiss, Inc.). The X-104 (hZO-2) gene was identified in relation to the FIG. 1. Schematic representation of 4.1R and X-104 (hZO-2) peptides found to interact in the yeast twohybrid screening of a human brain cDNA library. A, the organization of protein 4.1R (P 4.1R) and its 22-24-kDa domain was used as the bait in the yeast two-hybrid assays. Different exons of the 4.1R gene and corresponding chymotryptic domains are shown. The asterisks represent the alternatively spliced exons. B, shown is a schematic representation of the structural organization of the hZO-2 protein and its segments encompassed by the positive clones obtained from the yeast two-hybrid screen. GUK, guanylate kinase; ϩ, basic domain; Ϫ, acidic domain; ␣, alternatively spliced. Colocalization of 4.1R and ZO-2 in MDCK Cell Tight Junctions-To verify the interaction of 4.1R with ZO-2 and to locate the site(s) of their intracellular interaction, we examined the subcellular distribution of 4.1R and ZO-2 in confluent MDCK cells by double-label immunofluorescence microscopy. As shown in Fig. 2 (A and B), ZO-2 (red) and 4.1R (green) localized at the cell-cell contacts and displayed honeycomb-like staining patterns. Both ZO-2 and 4.1R also showed some diffuse cyto- Occludin has been shown to exclusively localize at TJs of epithelial and endothelial cells (52,53) in association with ZO-2 and ZO-1. To verify the localization of 4.1R at tight junctions, we performed double immunofluorescent staining of confluent MDCK cells with 4.1R and occludin. As shown in Fig. 2, occludin (red; panel D) and 4.1R (green; panel E) localized at apical cell borders and, to some extent, were diffuse in the cytoplasm. A transverse view of the cells taken at the position of the dotted line (in Fig. 2, D and E) showed that both occludin (Fig. 2DЈ) and 4.1R (Fig. 2EЈ) localized at the TJ along the apical side of the cells. As shown in Fig. 2F, 4.1R precisely colocalized with occludin at the MDCK cell TJs. Colocalization of 4.1R with ZO-2 and occludin at MDCK cell TJs was also observed when cells were stained with anti-10-kDa Ab or anti-16-kDa Ab (data not shown). The colocalization of 4.1R with occludin and ZO-2 indicated that 4.1R localizes specifically to TJs of confluent MDCK cells. To study the subcellular localization of ZO-2 and 4.1R in more detail, we performed high resolution immunogold electron microscopy using fixed confluent MDCK

cells. Anti-ZO-2 Ab was used as a control. As shown in Fig. 2 (G and H), labeling for ZO-2 (10-nm gold particles) and 4.1R (5-nm gold particles) was found concentrated at the tight junctions as clusters (arrows). This further supports our contention that 4.1R localizes at tight junctions. In Vivo Association of 4.1R and ZO-2-To examine the interaction of 4.1R with ZO-2 in vivo, we performed immunoprecipitation using confluent MDCK cell extracts. The extracts were subjected to immunoprecipitation with anti-HP Ab or anti-ZO-2 Ab. Preimmune serum (rabbit), purified rabbit IgG, and anti-p53 Ab (an irrelevant Ab) were used as controls. Analysis of immunoprecipitates by immunoblotting using anti-HP Ab showed that anti-ZO-2 Ab coprecipitated two polypeptides of Х135 and Х150 kDa; these comigrated with two polypeptides of similar molecular mass that were immunoprecipitated by anti-HP Ab (Fig. 3A). Similarly, analysis of the immunoprecipitates by immunoblotting using anti-ZO-2 Ab showed that ZO-2 coprecipitated with 4.1R (Fig. 3B). Neither 4.1R nor ZO-2 was coprecipitated with preimmune serum, rabbit IgG, or anti-p53 Ab (Fig. 3, A and B). These results suggest that 4.1R and ZO-2 are associated together in vivo. Analysis of the same immunoprecipitates by immunoblotting using anti-24-kDa Ab also revealed the presence of Х135and Х150-kDa polypeptides (Fig. 3C), confirming that these polypeptides are 4.1R isoforms. However, these isoforms of 4.1R were not detected in anti-HP Ab or anti-ZO-2 Ab supernatants by anti-24-kDa Ab, but were detected by anti-HP Ab. Because anti-24-kDa Ab is raised against exon 19, these data suggest that most of the 135-and 150-kDa 4.1R isoforms that contain alternative exon 19 coprecipitate with ZO-2 and therefore are not detected in the supernatant, whereas those isoforms not containing exon 19 do not coprecipitate with ZO-2 and thus are detected by anti-HP Ab. This is consistent with the results of yeast twohybrid mapping data that exons 19 -21 of 4.1R are required for interaction with ZO-2 (Fig. 4B). We determined the efficiencies of immunoprecipitation and coprecipitation to discern the fractions of 4.1R and ZO-2 that associate together. Quantitation from three different experiments (representative gels are shown in Fig. 3, A and B) showed that 80 -90% of 4.1R was precipitated by anti-HP Ab, which in turn coprecipitated 35-48% of ZO-2. In addition, Ͼ95% of ZO-2 was precipitated by anti-ZO-2 Ab, which coprecipitated 45-55% of 4.1R. Analysis of the post-extraction pellet (insoluble fraction) revealed that only 5-10% of both 4.1R and ZO-2 were not extracted (data not shown). This is consistent with the notion that both 4.1R and ZO-2 move out of the nucleus at confluent density in MDCK cells (75,76). Thus, ϳ35-45% of total cellular 4.1R and ZO-2 associate together in confluent MDCK cell TJs. Taken together, all of these results strongly suggest that about half of the cytoplasmic 135-kDa nonerythroid isoforms of 4.1R in MDCK cells interact with ZO-2. The complex splicing pattern of 4.1R generates distinct isoforms with insignificant differences in molecular mass, but with a significant effect on their subcellular targeting. Although the isoform(s) of 4.1R that interact with hZO-2 are also of the same higher molecular class (ϳ135 kDa) that interacts with NuMA in interphase nuclei and during mitosis (13), the same isoform(s) may not interact with NuMA because of their distinct subcellular localization. However, it has been shown that a given 4.1R isoform can adopt several localizations within the cell (11 3. 4.1R and ZO-2 coprecipitate in confluent MDCK cell extracts. MDCK extracts were prepared as described under "Experimental Procedures" and subjected to immunoprecipitation using different antibodies. The immunoprecipitates were analyzed by immunoblotting using anti-HP Ab (A), anti-ZO-2 Ab (B), and anti-24-kDa Ab (C). (In C, anti-ZO-2 Ab appears to coprecipitate more 4.1R than anti-HP Ab because of unequal loading.) One-fourth of the immunoprecipitates of anti-HP Ab, preimmune serum, anti-ZO-2 Ab, rabbit IgG, and anti-p53 Ab and one-eighth of the supernatant fractions of anti-HP Ab (anti-HP Ab sup.) and anti-ZO-2 Ab (anti-ZO-2 Ab sup.) immunoprecipitates were loaded in order. segments of 4.1R and hZO-2 or ZO-2 that interact, we examined the interactions between different domains or their segments of 4.1R and hZO-2 or ZO-2 in the yeast two-hybrid assays. We expressed different domains or segments of 4.1R as fusion products of the Gal4 DNA-binding domain in pAS2-1 and cotransformed them into Y190 with different segments of hZO-2 or ZO-2. The latter were expressed as Gal4-AD fusion products in pACT2, as shown in Fig. 4A. Full-length 4.1R (135 kDa), its 80-kDa isoform, and its 22-24-kDa domain interacted with 1) full-length hZO-2, 2) a segment of the C-terminal proline-rich domain of hZO-2 (aa 980 -1168), and 3) amino acids 187-1174 of ZO-2. They did not interact with the N-terminal part of ZO-2 (aa 1-189) (Fig. 4A). The N-terminal extension (HP) and the 30-, 16-, and 10-kDa domains of 4.1R failed to interact with hZO-2 or ZO-2. None of the domains of 4.1R, when expressed as Gal4-BD fusion proteins in Y190, expressed the reporter gene(s) by themselves or in combination with Gal4-AD in pACT2 alone (data not shown). These data suggest that the C-terminal domains of 4.1R and hZO-2 or ZO-2 are necessary and sufficient for their interaction. As illustrated in Fig. 4B, exons 17 and 18 were not required for interaction of hZO-2 with 4.1R. Fusion proteins consisting of exons 19 -21 of 4.1R (exons 19 -21/pGBT9) bound to hZO-2, but none of these exons alone (exon 19/pGBT9, exon 20/pGBT9, or exon 21/pGBT9) was sufficient for binding. Amino acids 980 -1168 were encoded by the hZO-2 clones obtained from the two-hybrid screening. As shown in Fig. 4C, the C-terminal truncation of the fusion protein from amino acids 1119 to 1168 did not impede binding, but further truncation from the Cterminal end (aa 1074 -1168) abolished the binding of hZO-2 to 4.1R. From the N-terminal end, truncation of the hZO-2 fusion protein from its N terminus up to amino acid 1053 did not affect and their segments were expressed as Gal4-BD or Gal4-AD fusion proteins, respectively, in yeast strain Y190 by cotransformation and were assayed for the expression of the reporter genes (see "Experimental Procedures" for details). Plus signs indicate the expression of the reporter genes lacZ and HIS3 (and thus the interaction between the peptides), and minus signs indicate non-expression of the reporter genes (and thus no interaction between the peptides). B, the C-terminal 96 amino acids of 4.1R are sufficient for its interaction with hZO-2. Schematic diagrams of the various exons within the 22-24-kDa domain of 4.1R (white rectangles) fused to the DNA-binding domain of Gal4 (shaded rectangles) used in the two-hybrid assay are shown. The names of the plasmids that encoded each construct are given to the right of each schematic diagram. Plasmid pAS2-1 expresses the Gal4 DNA-binding domain alone and was used as a negative control. These plasmids were cotransformed with hZO-2/pACT2 expressing amino acids 980 -1168 of hZO-2 fused to the activation domain of Gal4. C, amino acids 1054 -1118 of hZO-2 are sufficient for its interaction with 4.1R. Schematic diagrams of the various hZO-2 polypeptides fused to the activation domain of Gal4 (hatched) used in the two-hybrid assay are shown. Plasmid pACT2, which expresses the Gal4 activation domain alone, was used as a negative control (data not shown). These plasmids were cotransformed with pGBT9 or pAS2-1 expressing the C-terminal domain of 4.1R or its exons 19 -21 fused to Gal4-BD. ␤-gal, ␤-galactosidase; GUK, guanylate kinase; Alt. Spl., alternatively spliced. its binding to 4.1R. The fusion protein encoding amino acids 1054 -1118 of hZO-2 bound to 4.1R. The binding strength was comparable to the fusion protein encoding amino acids 980 -1168 of hZO-2, as assessed by expression of the reporter gene lacZ (data not shown). Therefore, it appears that amino acids 1054 -1118 of hZO-2 are sufficient for its interaction with 4.1R. These results suggest that the amino acids through which 4.1R and hZO-2 interact reside within amino acids encoded by exons 19 -21 of 4.1R and amino acids 1054 -1118 of hZO-2. Protein 4.1R Associates with Tight Junction Proteins ZO-1, ZO-2, and Occludin in Confluent MDCK Cells-To examine the association of 4.1R with the TJ protein complex, we asked if it occurred in association with other TJ components such as ZO-1 and occludin (54 -56). Therefore, we performed immunoprecipitation using confluent MDCK cell extracts and looked for 4.1R in the immunoprecipitates of ZO-1 and occludin and vice versa. Rabbit preimmune serum, purified rabbit IgG, and anti-p53 Ab (an irrelevant Ab) were used as negative controls to rule out nonspecific aggregation. ZO-2, which is known to coprecipitate with and bind directly to ZO-1 (57,58) and occludin (41,58), was used as a positive control. As stated earlier and shown in Fig. 5A, two 4.1R isoforms (Х135 and Х150 kDa) were immunoprecipitated by anti-HP Ab. These isoforms of 4.1R were also found to coprecipitate with ZO-2, ZO-1, and occludin when immunoprecipitation was carried out with anti-ZO-2 Ab, anti-ZO-1 Ab, and anti-occludin Ab, respectively, but not with preimmune serum, rabbit IgG, or anti-p53. Analysis of the same immunoprecipitates by immunoblotting using antibodies specific for the 16-kDa domain of 4.1R (anti-16-kDa Ab) also revealed the ϳ135and ϳ150-kDa 4.1R isoforms in the immunoprecipitates of anti-HP Ab, anti-ZO-2 Ab, anti-ZO-1 Ab, and anti-occludin Ab. No other 4.1R isoforms were seen in the immunoprecipitates of ZO-2, ZO-1, and occludin, suggesting that the isoforms of 4.1R that associate with these TJ proteins are of the 135-kDa molecular mass class. This study also confirmed the Х150-kDa protein as a 4.1R isoform. To examine the involvement of other 4.1R-binding cytoskeletal proteins such as actin and spectrin in this complex, we analyzed the anti-HP Ab, anti-ZO-2 Ab, anti-ZO-1 Ab, and anti-occludin Ab immunoprecipitates by immunoblotting using the relevant antibodies. As shown in Fig. 5F, actin was found to coprecipitate with 4.1R, ZO-2, ZO-1, and occludin, but not with p53, preimmune serum, or rabbit IgG immunoprecipitates. Anti-spectrin Ab detected both the ␣and ␤-isoforms of spectrin in MDCK cell lysates (Fig. 5G). Interestingly, only ␣-spectrin was detected in the anti-HP Ab, anti-ZO-2 Ab, anti-ZO-1 Ab, and anti-occludin Ab immunoprecipitates. The absence of actin and spectrin (despite their abundance in the cell) in immunoprecipitates of preimmune serum, rabbit IgG, and anti-p53 Ab indicated that their presence in the immunoprecipitates of anti-HP Ab, anti-ZO-2 Ab, anti-ZO-1 Ab, and anti-occludin Ab was due to a specific intracellular association with these proteins. The coprecipitation of actin and ␣-spectrin with the TJ proteins was not surprising because the C-terminal half of ZO-1 has been shown to cosediment with actin filaments (58,59), and ␣-spectrin has been shown to coprecipitate with ZO-1 (60, 61). To examine the presence of other 4.1R-binding and/or TJassociated proteins in this protein complex, anti-HP Ab immunoprecipitates were examined by SDS-polyacrylamide gel electrophoresis and silver staining (Fig. 5H). About 17 protein bands were visualized in the anti-HP Ab immunoprecipitates, but not in the preimmune serum immunoprecipitates. The molecular mass of some of these proteins was identical to that of some of the TJ-associated proteins. A duplicate of the gel in Fig. 5H was transferred to polyvinylidene difluoride membrane and analyzed by immunoblotting for the identification of 4.1R, ZO-1, ZO-2, occludin, actin, and spectrin using the relevant antibodies (data not shown). The protein bands on the silverstained gel that were identified by immunoblotting are labeled (Fig. 5H). These results suggest that 4.1R isoform(s) and TJ proteins ZO-1, ZO-2, and occludin (and possibly others) along with actin and ␣-spectrin associate together in vivo. Protein 4.1R Does Not Associate with Tight Junction Proteins ZO-1, ZO-2, and Occludin in Nonconfluent MDCK Cells-To examine whether or not the association of 4.1R isoforms is TJ-specific, we repeated the immunoprecipitation experiments in nonconfluent (Ͻ50% confluent) MDCK cell lysates (TJ not organized). As shown in Fig. 6A, in contrast to confluent cell lysates, only one isoform of 4.1R (Х135 kDa) was precipitated by anti-HP Ab. Because 135-kDa 4.1R was expressed in confluent as well as nonconfluent cells, but ϳ150-kDa 4.1R was detected in confluent cells only, it appears that ϳ150-kDa 4.1 might play a critical role in tight junction organization. However, our initial efforts to delineate the proper exon combination of the ϳ150-kDa isoform have not yielded precise identity of this isoform. The characterization of the ϳ150-kDa isoform is in progress. Anti-ZO-1 Ab, anti-ZO-2 Ab, and anti-occludin Ab failed to co-immunoprecipitate 4.1R in the nonconfluent state (Fig. 6A), even though ZO-1 was coprecipitated with ZO-2 and vice

versa (Fig. 6, B and C). The fractions of ZO-2 and ZO-1 that coprecipitated were also less compared with those of the confluent cell lysates. Occludin also did not coprecipitate with 4.1R, ZO-1, ZO-2, or any of the controls (Fig. 6D). Analysis of the immunoprecipitates by immunoblotting showed that neither actin (Fig. 6E) nor spectrin (Fig. 6F) was coprecipitated with 4.1R, ZO-1, ZO-2, or occludin. The difference in results observed in contrast to confluent cells was not due to nonexpression of the above proteins because all these proteins were detected in whole cell lysate (Fig. 6, A-F). The results from Figs. 5 and 6 suggest that the association of 4.1R, ZO-1, ZO-2, occludin, actin, and ␣-spectrin together in MDCK cells is dependent on TJ formation, which is organized only when these cells become confluent. Protein 4.1R Binds to ZO-2, ZO-1, and Occludin in Vitro-Because the results from immunoprecipitation experiments suggested that 4.1R associates with TJ proteins ZO-2, ZO-1, and occludin in vivo, we asked if there were additional binary interactions between 4.1R and ZO-1 or occludin. We constructed yeast expression vectors expressing full-length ZO-1, occludin, or their different segments in frame with Gal4-AD and performed yeast two-hybrid assays with 4.1R and its subdomains. The results of filter lift ␤-galactosidase activity assays were inconclusive. Therefore, chlorophenol red-␤-D-galactopyranoside-based liquid ␤-galactosidase activity assays were performed. As shown in Fig. 7A, the N-terminal cytoplasmic domain of occludin (aa 1-57) did not show substantial ␤-galactosidase activity over the control. However, full-length occludin (data not shown) as well as its C-terminal cytoplasmic domain (aa 250 -504) showed weak interactions with the 135-and 80-kDa isoforms and C-terminal domain of 4.1R. A weak interaction was also observed between the N-terminal cytoplasmic domain of occludin (aa 1-57) and HP of 4.1R. In addition, weak interactions were detected between the 135-and 80-kDa isoforms and the 24-kDa domain of 4.1R and the C-terminal proline-rich domain (aa 1045-1737) (Fig. 7B), but not the Nterminal part (aa 1-1044) of ZO-1 (data not shown). To confirm direct binding of 4.1R to ZO-2, ZO-1, or occludin, we performed in vitro binding assays. Different isoforms and domains of 4.1R were expressed as GST fusion proteins and purified by coupling to GST-Sepharose beads as described (13). revealed, respectively. These results suggest direct binding of ZO-2, ZO-1, and occludin to 4.1R and are consistent with results from the yeast two-hybrid assays. Recruitment of 4.1R/GFP Fusion Proteins to Tight Junctions in Confluent MDCK Cells-To verify that 4.1R interacts with ZO-2 and/or other tight junction-associated proteins in intact cells, we constructed expressions vectors with segments of 4.1R fused to GFP and transfected them into cultured MDCK cells. Because results from yeast two-hybrid and in vitro binding assays suggest that the 22-24-kDa domain of 4.1R is sufficient for interaction with ZO-2, ZO-1, and occludin, cells were transfected with 24-kDa/GFP. GFP alone and 10-kDa/GFP were used as controls. However, as shown in Fig. 9, GFP alone, 10-kDa/GFP, or 24-kDa/GFP was not recruited to cell-cell junctions (Fig. 9, A-C), but 10ϩ24-kDa/GFP was efficiently recruited to cell-cell junctions (Fig. 9, D and E). The state of confluency and localization of ZO-1 at cell-cell junctions in the same cells were revealed by staining the cells with anti-ZO-1 Ab (Fig. 9, AЈ-EЈ). Confocal microscopy also revealed that the 10ϩ24-kDa/GFP fusion proteins colocalized with ZO-2 and ZO-1 in cultured MDCK cell tight junctions (data not shown). All these fusion proteins were also concentrated in the nucleus. DISCUSSION Several studies suggest that the cortical cytoskeleton is involved in the structural and functional organization of TJs (reviewed in Ref. 56; Ref. 62), but little is known about the molecule(s) that link these two distinct structures. In this study, we demonstrate that two nonerythroid isoforms of 4.1R (ϳ135 and ϳ150 kDa) interact with hZO-2. In addition to ZO-2, these isoforms of 4.1R also associate with other TJ proteins such as ZO-1 and occludin and the cytoskeletal proteins ␣-spectrin and actin in one protein complex. Endogenous components of TJs also efficiently recruit tagged 4.1R segments containing the spectrin/actin-and ZO-2-binding domains to TJs. We thus hypothesize that 4.1R isoforms in epithelial cells may participate in regulation of TJ by mediating a direct link between the TJ proteins with the underlying cytoskeleton. ZO-2 belongs to the MAGUK protein family. MAGUK proteins such as p55, human DLG, and the human LIN2 homolog are known to associate with the cortical actin cytoskeleton. In erythrocytes, the MAGUK protein p55 links the transmembrane protein glycophorin C to the spectrin/actin cytoskeleton through the 30-kDa domain of protein 4.1R (20). Human DLG and the human LIN2 homolog are also known to bind to 4.1R in epithelial cells (15,29). Similarly, ZO-1, a member of the MAGUK family of proteins, binds to ZAK, a serine-threonine kinase (63), and links the transmembrane protein occludin to cytoskeleton by binding to F-actin (59). However, the interaction between hZO-2 and 4.1R was unexpected because hZO-2 lacks the 4.1R-binding motif used by the MAGUK proteins p55, human DLG, and human LIN2, a conserved lysine-rich sequence motif located between the SH3 and guanylate kinase domains (15,20,29). Although the C-terminal domain encoded by exons 19 -21 of 4.1R interacts with multiple PDZ domains containing hZO-2, the expected PDZ domain-tail interaction (64) was not ob- Because deletion of exon 19, 20, or 21 abolishes this interaction, it appears that the amino acids of 4.1R that interact with hZO-2 may be distributed in these three exons and that there may be multiple contact sites between these proteins. It is also possible that the presence of these three exons of 4.1R together gives rise to a particular folding of 4.1R required for its interaction with hZO-2 or ZO-2. A recent study shows that these amino acids of 4.1R are highly conserved among other 4.1R-like genes (30), raising the possibility that these gene products may also interact with hZO-2 and ZO-2. ZO-2 is known to associate with the cytoplasmic surface of the TJ (51,57). It shares a strong homology with ZO-1, especially within the conserved MAGUK domains (39). It is expressed in almost every tissue (44) and associates with ZO-1 through its second PDZ domain at both the adherens junctions and TJs (41,51,58,65). ZO-2 binds directly to the C-terminal 147 amino acids of occludin (66) and coprecipitates with occludin and ␣-catenin (41). It is a component of the TJ along with ZO-1, ZO-3, cingulin, 7H6 antigen, symplekin, several unidentified proteins, and transmembrane proteins such as occludin (reviewed in Ref. 56) and members of the claudin family (67). ZO-1, ZO-2, and occludin have been shown to localize at TJs of confluent MDCK cells (51-53, 61). Using anti-HP Ab, an antibody specific for nonerythroid isoforms of 4.1R, we found that 4.1R epitopes are also localized at TJs of MDCK cells by confocal microscopy and high resolution electron microscopy (Fig. 2, B and H, respectively). Confocal microscopic analysis of double-labeled confluent MDCK cells showed that these 4.1R epitopes colocalize at TJs with ZO-2 and occludin (Fig. 2, C and F). In immunoprecipitation studies using the same antibodies, ZO-2 and two nonerythroid 4.1R isoforms were found to coprecipitate together. About 40 -50% of 4.1R and ZO-2 appear to remain associated together under our experimental conditions. Unlike ZO-2 and ZO-1, which also localize at adherens junction in addition to TJs, occludin is exclusively localized at the TJs (52,53). We found that 4.1R, ZO-1, ZO-2, and occludin coprecipitated together. This association of 4.1R with ZO-2, ZO-1, or occludin was not seen in cells that were not confluent (and thus in which the tight junction was not organized) (Fig. 6), strongly suggesting that these proteins associate together at TJs. Coprecipitation does not reveal whether there is a direct interaction between 4.1R and ZO-1 or occludin. We thus tested this hypothesis in yeast two-hybrid and in vitro binding assays. As shown in Fig. 7, only weak interactions between the Cterminal domain of 4.1R and the proline-rich domain of ZO-1 (Fig. 7B) or the C-terminal cytoplasmic domain of occludin ( Fig. 7A) were observed in yeast two-hybrid assays. These interactions were only 25-30% of the interaction between 4.1R and hZO-2 or ZO-2 as detected in yeast two-hybrid assays. ZO-2, ZO-1, and occludin bound to 4.1R in in vitro binding assays (Fig. 8). Interestingly, a recent study suggested that a third protein might mediate the interaction between ZO-1 and ZO-2 (58). Similarly, a mutant occludin that bound to ZO-1 in vitro failed to localize at TJs (68). Both ZO-1 and ZO-2 also localize to TJs in occludin-deficient TJs (41,69). These findings imply that association of other factors, in addition to ZO-1 and ZO-2, may be required for localization of occludin at TJs. Both the 135-and 80-kDa isoforms interact with ZO-2 in yeast two-hybrid assays, whereas HP alone does not. Thus, our results do not define the precise isoform(s) of 4.1R that interact with ZO-2. Because of the complex splicing pathway that 4.1R undergoes, it is difficult to know exactly which exons are included in the isoform(s) that interact with ZO-2. However, epitopes for anti-HP Ab are observed at TJs; moreover, the two isoforms of 4.1R that coprecipitate with ZO-2 and other TJ proteins are exclusively of the higher molecular mass. Finally, no 80-kDa isoform containing exon 19 (which is required for its interaction with ZO-2) is expressed in MDCK cells (13). We thus believe that isoforms of 4.1R that interact with ZO-2 are of the 135-kDa class or of similar molecular mass. The 80-kDa isoform binds to ZO-2 in yeast two-hybrid and in vitro binding assays, but does not coprecipitate with ZO-2. These data are compatible because the increased concentration and proximity of molecules containing the binding sites in yeast two-hybrid and in vitro binding assays could allow the 80-kDa isoform to interact with ZO-2. In the intact cells, factors such as posttranslational modifications and compartmentalization or presence of modifying cofactors will affect their interaction. In the same context, the amino acids encoded by exons 19 -21 of 4.1R are required and sufficient for interaction of 4.1R with ZO-2 in vitro, but not in vivo. Indeed, the 22-24-kDa domain is not sufficient to target GFP to tight junctions, although per se it binds well to ZO-2. These amino acids are highly conserved among the 4.1-like gene family (30). It can thus be argued that other 4.1R-like gene products interact with ZO-2. However, as discussed above, our results strongly implicate 4.1R isoforms as opposed to 4.1-like gene products because the anti-HP Ab used in our assays does not react with other 4.1-like proteins. A noteworthy finding is the association of 4.1R with ZO-1, FIG. 9. Localization of transiently expressed 4.1R/GFP fusion proteins in confluent MDCK cells. GFP or the 10-kDa/GFP, 24-kDa/ GFP, or 10ϩ24-kDa/GFP fusion proteins were exogenously and transiently expressed in MDCK cells. The cells were grown to confluence, fixed, and stained with anti-ZO-1 Ab (in red; AЈ-EЈ) for localization of ZO-1. Expression of GFP or GFP fusion proteins was visualized by green fluorescence (A-E). 10ϩ24-kDa/GFP fusion proteins were recruited to cell-cell junctions, but not GFP alone, 10-kDa/GFP, or 24-kDa/GFP. Anti-ZO-1 antibody-stained cells that correspond to cells transfected with GFP or 4.1R/GFP fusion proteins are indicated with arrows in corresponding pictures. Magnification ϫ 40. ZO-2, occludin, ␣-spectrin, and actin apparently in one protein complex, implying a connection of TJs to the cortical cytoskeleton, as previously suggested (Refs. 62 and 70; reviewed in Ref. 56). Several studies also suggest that actin plays a role in regulation of TJ permeability (71,72). ZO-1 has been shown to cosediment with F-actin (58,59) and to colocalize with actin (60,(72)(73)(74). ZO-1, ZO-2, and occludin cosediment with ␣-spectrin (68). ZO-1 has been suggested to link the actin cytoskeleton to TJ via an "actin/ZO-1/occludin" linkage (58,59). The Cterminal parts of ZO-2 and ZO-3 show poor homology to the C-terminal half of ZO-1, which cosediments with actin. ZO-2 and ZO-3 also lack the lysine-rich "protein 4.1-binding motif" found in other MAGUK proteins. These observations suggest that ZO-1 may be the only MAGUK protein at the TJ that can bind to actin and thereby serve as the structural and signaling component of TJs. In contrast, a recent in vitro study suggests that ZO-1, ZO-2, and occludin can directly interact with F-actin (66). The following evidence from our study suggests that 4.1R may provide or supplement the linkage between the TJ and the cortical cytoskeleton. First, ZO-2, ZO-1, and occludin were

found to interact with 4.1R, which in turn is known to bind to the cytoskeletal proteins actin and spectrin. Second, the association of actin and ␣-spectrin with ZO-1, ZO-2, or occludin was not observed in nonconfluent cells when 4.1R also failed to interact with these proteins. Third, ZO-2, ZO-1, and occludin interacted with 4.1R in in vitro binding and yeast two-hybrid assays. Finally, GFP-tagged segments of 4.1R that contained both the spectrin/actin-and ZO-2-binding domains were recruited to TJs, but the segments that contained only the spectrin/actin-binding domain or the ZO-2 binding domain were not. Thus, as shown in Fig. 10, ZO-2, like other MAGUKs, may be establishing a link between the TJ and the actin-based cytoskeleton via a different binding interaction with 4.1R. This function of 4.1R in nonerythroid cells is not established, but unpublished data from our laboratory suggest that the 135-kDa nonerythroid isoform of 4.1R forms a ternary complex with F-actin and fodrin (nonerythroid spectrin) in vitro. 2 Interaction of 4.1R with members of the MAGUK family, ZO-1, and ZO-2 could assist in recruitment of other proteins to coordinate cell signaling. Other members of the protein 4.1 superfamily such as merlin, ezrin, radixin, and moesin are known to be involved in different signaling pathways (36,37). It is thus tempting to speculate that 4.1R may have a role in signal transduction between the TJ and the cytoskeleton. A similar role of 4.1R has been suggested for the human CASK and 4.1R interaction (29). Risk factors for acute kidney injury and mortality in high risk patients undergoing cardiac surgery Background Acute Kidney Injury (AKI) represents a clinical condition with poor prognosis. The incidence of AKI in hospitalized patients was about 22–57%. Patients undergoing cardiac surgery (CS) are particularly exposed to AKI because of the related oxidative stress, inflammation and ischemia-reperfusion damage. Hence, the risk profile of patients undergoing CS who develop AKI and who are consequently at increased mortality risk deserves further investigation. Methods We designed a retrospective study examining consecutive patients undergoing any type of open-heart surgery from January to December 2018. Patients with a history of AKI were excluded. AKI was diagnosed according to KDIGO criteria. Univariate associations between clinical variables and AKI were tested using logistic regression analysis. Variable thresholds maximizing the association with AKI were measured with the Youden index. Multivariable logistic regression analysis was performed to assess predictors of AKI through backward selection. Mortality risk factors were assessed through the Cox proportional hazard model. Results We studied 158 patients (mean age 51.2±9.7 years) of which 74.7% were males. Types of procedures performed were: isolated coronary artery bypass (CABG, 50.6%), valve (28.5%), aortic (3.2%) and combined (17.7%) surgery. Overall, incidence of AKI was 34.2%. At multivariable analysis, young age (p = 0.016), low blood glucose levels (p = 0.028), estimated Glomerular Filtration Rate (p = 0.007), pH (p = 0.008), type of intervention (p = 0.031), prolonged extracorporeal circulation (ECC, p = 0.028) and cross-clamp (p = 0.021) times were associated with AKI. The threshold for detecting AKI were 91 and 51 minutes for ECC and cross-clamp times, respectively. At survival analysis, the presence of AKI, prolonged ECC and cross-clamp times, and low blood glucose levels forecasted mortality. Conclusions AKI is common among CS patients and associates with shortened life-expectancy. Several pre-operative and intra-operative predictors are associated with AKI and future mortality. Future studies, aiming at improving prognosis in high-risk patients, by a stricter control of these factors, are awaited. Introduction Acute Kidney Injury (AKI) represents a severe clinical condition characterized by an increased risk of mortality, particularly in hospitalized patients [1]. The global incidence of AKI in hospital settings, according to Kidney Disease: Improving Global Outcomes (KDIGO) definitions is about 22% -57%, with an increasing prevalence due to the concomitant increasing trend in cardiovascular disease comorbidity [2,3]. The onset of AKI is common among patients undergoing cardiac surgery (CS) because of a series of intrinsic factors, including the transfusion of large volumes of exogenous blood products, extracorporeal circulation techniques, high doses of vasopressors, which, all together, determine cycles of ischemia-reperfusion, oxidative damage, and inflammation [4,5]. The incidence of AKI after CS is related to the type of operative procedure [6]. In a large cohort study, enrolling more than 3.000 patients who underwent CS and without a previous history of kidney disease, the rate of AKI was higher (reaching up to 59% patients) in aortic surgery, including aortic interventions combined with other interventions, as compared to isolated coronary artery bypass surgery (CABG) (37%), valve surgery (49%) and thoracic surgery (33%) [7]. Moreover, patients who developed AKI were also at significant increased risk of long-term mortality if they underwent CABG, aortic and thoracic surgery, but not valve surgery [7]. However, this topic is still controversial. Another observational analysis, indeed, found no difference in 3-month mortality risk among patients with AKI, who underwent different types of Cardiac Surgery [8]. Owing to these controversies, the knowledge of risk factors for AKI remains a crucial research endeavor, since it could help clinicians in improving prevention, a prompt diagnosis and management of AKI in this setting of patients. Several risk scores for AKI have already been published [9,10]. Among them, the 'any-stage AKI risk score' has shown a good performance for all stages of AKI and is also available in clinical practice as a web-based calculator [10]. Nevertheless, the inclusion of novel risk factors of AKI in a risk prognostic model has been widely prompted, also on the basis of the results of several clinical trials that have failed in demonstrating the efficacy of prevention strategies, such as the use of statins at high-doses or the remote ischemic preconditioning (RIPC) treatment [4,11,12]. We designed a retrospective, observational, study enrolling patients who underwent CS to evaluate both the predictors of the onset of AKI and the role of cardiorenal markers on individual prognosis (i.e. mortality endpoint) over time. The study was approved by the Institutional Review Board (IRB) of CIFL -Interuniversity Center of Phlebolymphology-at Magna Graecia University of Catanzaro (Id approval number: ER. ALL.2018.37A). All the individual data collected were fully anonymized before the analysis. Patients were enrolled according to the following criteria: age > 18 years, undergoing CS with extracorporeal circulation (ECC). Patients requiring emergency operation, with a preoperative infective status or with a history of AKI during the previous 3-months, or renal replacement therapy were excluded. AKI was defined according to the KDIGO classification, based on the change in serum creatinine levels (� 0.3 mg/dL within 48 hours or � 50% within 7 days) or urine output volume less than 0.5 mL/kg/hour for >6 hours [13]. Surgical procedures were realized through median sternotomy. In CABG, the left internal mammary artery (LIMA) was harvested in a pedicled fashion and anastomosed to the left anterior descending coronary artery (LAD). When harvested, a pedicled radial artery (RA) was used as Y-graft with LIMA. Single or sequential saphenous vein grafting was left to the surgeon's discretion. Valvular procedures included aortic (AVR) and mitral (MVR) valve replacement. An age of 65 years was used as the threshold for choosing mechanical or biological prosthesis after counseling with the patient. Surgical procedures across the aorta included prosthetic replacement of ascending aorta in patients with aneurysmatic dilation. Aortic clamping was performed in all operations. The use of ECC was standardized. Total ECC flow was maintained at 2.6 L/min/m 2 in all operations. Systemic temperature was kept at 34˚C. Myocardial protection was always achieved with intermittent antegrade and retrograde hyperkalemic blood cardioplegia [14][15][16][17][18]. Preoperative data about comorbidities, blood pressure, body mass index (BMI) and therapy were collected. Laboratory values, echocardiographic measures and arterial blood gas (ABG) were obtained at 5 different times: upon hospital admission, which was considered the baseline study visit; at 1h, 24h and 48h after the operation; at discharge. For hemoglobin (hgb) <7.5 g/ dL (or hematocrit [Hct] <22 percent), initial treatment during CPB is removal of fluid by hemoconcentration when possible. Transfusion of packed red blood cells (RBCs) is reasonable if Hgb remains <7.5 g/dL when ultrafiltration is not possible or is ineffective [19][20][21][22][23][24]. The presence of hypoglycemia, as a categorical variable, was defined by the presence of blood glucose < 75 mg/dL. Postoperative data regarding the main clinical parameters and medications were also collected. After the hospital discharge, patients were followed for monitoring complications and treatment with periodical in-hospital visits with a frequency variable based on the severity of their clinical parameters. Follow-up lasted until August 31 st 2020, the onset of death or the last visit to the CS Unit. Statistical analysis Continuous variables were reported as mean ± standard deviation (SD) or median and interquartile range (IQR) based on their distribution. Comparison between groups was assessed by unpaired t-test or Mann-Whitney test. Categorical variables were analyzed using the Chisquare test. Univariate association between the main clinical variables and the onset of AKI was assessed by means of logistic regression analysis. To find a cut-point that maximizes the variable's ability to differentiate AKI from no-AKI endpoint the Youden index (J) was computed. Next, a backward variable selection method, with an elimination criterion of p<0.10, was performed to fit the multivariable logistic regression model. Multicollinearity was assessed with variance inflation factors (VIF), with values greater than 10 considered as cause for concern. First-order interactions between covariates included in the models have been tested. For the survival analysis, multivariable Cox proportional hazard regression analysis was used by assessing the effect of clinical or laboratory parameters on the onset of mortality over time. Data was analyzed using STATA 14 (StataCorp. College Station, TX, USA). Results Two out 160 patients were excluded from the initial cohort because of in-hospital death. The overall cohort (Table 1) was characterized by a high-risk profile as testified by the high frequency of diabetes (40.5%) and arterial hypertension (79.1%) as well as by a high BMI (27.4 ±4.3 kg/m 2 ). In the overall cohort, isolated CABG was the most performed operation (Table 2). Furthermore, in the whole cohort, AKI occurred in 54 out 158 patients (34.2%). AKI group was characterized by a statistically significant younger age (p = 0.040), a lower pre-operative eGFR (p<0.001), and higher extracorporeal circulation (ECC) and cross-clamp times (p = 0.036 and 0.053, respectively). Frequency of isolated CABG was higher in the non-AKI group, whereas other operative procedures were more prevalent in the AKI group (p = 0.014). The onset of AKI was also associated with an increased length of hospitalization, being of 3 days in median higher as compared with no-AKI group (p = 0.001). In the post-operative period, 53.2% patients were started with furosemide, 36.7% with amiodarone, 44.9% with norepinephrine and 5.1% with epinephrine treatment. Differences for these drugs frequencies (Table 3), a backward selection analysis was applied. The resulting multivariable logistic analysis, depicted in Table 4 showed that young age, lower blood glucose levels and eGFR, and pH reduction were associated with an increased risk for AKI. Similarly, prolonged ECC and cross-clamp times were found to be significant risk factors for AKI. Risk increased every 10 minutes of prolonged ECC (9%) and cross-clamp (13%) times. When Youden indexes have been computed, a value of 91 minutes (Sensitivity = 83% and Specificity = 78%) was found as the threshold of ECC that discriminates the risk for AKI, whereas the better discrimination for cross-clamp time was found above 51 minutes (Sensitivity = 85% and Specificity = 81%). When blood glucose was added as a categorical variable, the presence of hypoglycemia remained a strong and significant predictor of AKI (OR 2.55, 95% CI: 1.06-6.15). ECC and cross-clamp times have been separately added to the multivariable logistic regression due to collinearity (VIF = 24.48). Risk for AKI also varied by the type of Fig 1) with a 3.69-fold higher annual rate of mortality in AKI group. At multivariable Cox-proportional hazard model, adjusted for the covariates which differed as averaged values between death vs. no-death group (Fig 2 and Table 5) and the presence/ absence of AKI, lower blood glucose levels, ECC and cross-clamp times, and the presence of AKI were significant predictors for mortality endpoint. For sensitivity analysis, we replaced blood glucose level as a continuous variable with the presence/absence of hypoglycemia which was associated with an increased risk for all-cause mortality (HR 2.45, 95% CI: 1.35-4.43). Discussion Risk stratification for the cardiac surgical patient represents

an important tool that physicians may use for pre-operative decision-making and intra-operative management. In most cases, several comorbidities often affect this cohort of patients worsening the outcomes [7,8]. This evidence is confirmed in our cohort in which about 41% of patients suffered from type-2 diabetes, 79% from hypertension, and all patients (100%) were affected by previous cardiovascular disease. In addition, more than 30% of patients developed AKI or died during follow-up. It has been widely demonstrated that cardiovascular disease is strictly linked to kidney impairment, since an acute or chronic disorder in one organ may induce an acute or chronic disorder in another one [25,26]. The presence of cardiac dysfunction leads to the activation of the reninangiotensin-aldosterone-system (RAAS), sympathetic nervous system and vasopressin secretion, leading to fluid retention [27]. Moreover, CS procedures contribute per se to trigger AKI and, consequently to increase the individuals' risk [28]. The previous risk scores showed that a large number of preoperative, intraoperative and post-operative risk factors can predict the onset of AKI in patients undergoing CS [4,9,10,29,30]. However, these scores were not largely used in clinical practice, so that the scientific community prompted to implement studies exploring the risk factors of AKI and prognosis in high-risk patients, those who would most benefit from novel treatments and prevention strategies [31]. The methodological novelty of our study is the inclusion of both preoperative and intraoperative factors at the same time for evaluating the risk for AKI and subsequent mortality, with an adequate sample size. As the main results of this study, we confirmed the importance of well-established risk factors of AKI, such as the preoperative levels of eGFR, time of ECC and aortic cross-clamp times and expanded this association to other potential, still less explored, risk factors, such as the type of surgical procedure, blood glucose and blood pH. Furthermore, we showed that the risk factors for AKI may also predict a poor prognosis, besides and beyond the presence of AKI itself. With respect to AKI, ECC and cross-clamp times are two peculiar characteristics of CS that could elicit AKI [32]. Extracorporeal circulation contributes to AKI because of red blood cell hemolysis and the systemic inflammatory response syndrome (SIRS), that induces the increase in blood levels of mediators, such as IL-6, IL-8 and TNF-α [32][33][34]. A crucial point that influences these pathways is represented by the ECC duration. We observed that mean ECC time was slightly higher compared with that reported in other studies enrolling patients undergoing no-valve interventions, but it was comparable with the one measured in more complex surgical procedures [35,36]. However, those studies that evaluated ECC duration in both valve and no-valve intervention were limited by small sample sizes with a consequent weak power for statistical analysis [35,36]. We found that 91 minutes of ECC was found as the threshold above which the risk for AKI started to increase significantly. This evidence is higher when compared to an ECC time longer than 70 minutes reported by other studies [35][36][37]. Crossclamp time is another risk factor for renal dysfunction since it associates with ischemic-reperfusion damage. Cross-clamp time observed in our cohort was similar to that reported in other studies [35,36]. We found that 51 minutes of cross-clamp time was the best value which discriminates patients with and without AKI. The importance of ECC and cross-clamp times is even reinforced by the significant association of both parameters with an increased risk for mortality in our patients. We observed that patients, who had undergone valve, aortic and combined interventions, had about a 2-fold increased risk for AKI compared with those receiving isolated CABG. We could explain this finding by considering that isolated CABG is usually associated with lower ECC and cross-clamp times [38]. Moreover, patients undergoing procedures other than CABG are, per se, at higher basal risk. Interestingly, we found an interaction between ECC time and type of intervention (p = 0.001) as well as between cross-clamp time and type of intervention (p<0.001) with the onset of AKI. This may be explained with the higher ECC and cross-clamp times (132±48 minutes and 101±32 minutes, respectively) in patients with other procedures when compared to isolated CABG (115±44 minutes and 56±17 minutes, respectively). With respect to preoperative risk factors, we found that baseline eGFR was a strong determinant for the onset of AKI. This has been previously shown and has been recognized as one of the stronger predictors of AKI [7,34]. Originally, we found that young age and lower blood glucose levels can represent independent risk factors for AKI, regardless of the baseline levels of kidney function. With respect to the association between young age and AKI, this piece of data is apparently in contrast to previous literature reports which show how older age is associated with a raised risk of AKI [29,30]. However, we could explain this finding taking into account the inverse association between age and the systemic inflammatory response syndrome (SIRS) to the ECC. Previous studies have shown that the mean age of patients with SIRS is significantly lower than the one of those without SIRS [39,40]. Hence, younger patients seem to require a strong monitoring of renal function to the same extent as in older patients, because the risk for AKI remains non-trivial. Regarding glycemia, we may argue that almost all existing risk prediction models, evaluating the preoperative risk factors for AKI, are focused on the presence of diabetes as a determinant of AKI [9,10,29,30]. Based on our results, we could not exclude that also the opposite clinical condition, namely lower blood glucose levels, may be a predisposing factor for AKI. Indeed, evidence for an association between hypoglycemia and AKI have been recently shown, particularly in hospitalized patients [41]. Kidneys are directly involved in glucose metabolism, with the renal cortex being responsible for up to 30% of total gluconeogenesis. Hospitalization per se could enhance mechanisms that lead to hypoglycemia, such as inadequate nutrition intake, lack of counter regulatory mechanisms and reduction of kidney function [7,32,42,43]. Furthermore, we excluded that the association between AKI and blood glucose levels may be dependent of insulin treatment. In fact, frequency of insulin use before intervention did not differ (0 = 0.623) between patients who developed AKI and those who did not. We found that the presence of AKI itself is associated with an increased mortality risk over time and that this is partially explained by the complexity of surgical procedures as testified by the association of time of ECC and cross-clamp time with mortality rate. Interestingly, we found new evidence of association between AKI and mortality, by reporting an extended follow-up, reaching more than 20 months. This confirms the importance for clinicians to monitor kidney parameters which play a crucial prognostic role in high-risk patients [43][44][45][46][47]. Moreover, reduction in arterial pH levels are pivotal to diagnose acidosis, which is the most frequent in patients affected by renal impairment [48]. Metabolic acidosis is a predictor of AKI and deserves a strict monitoring since this parameter, if associated with the presence of hyperkalemia, oliguria, uremia and/or volume overload, represents an indicator for renal replacement therapy [49]. Strengths and limitations The single-center dimension of the present study limits the generalizability of the obtained results. Moreover, the small sample size did not allow to adjust for all the potential confounders of the association between CS and AKI. However, the literature lacks any strong evidence on this topic. The few papers investigating intraoperative variables studied more limited cohorts with a weaker statistical analysis compared with the present paper [36,37]. We did not collect glycated hemoglobin (HbA1c) values in our patients that would be useful to further investigate the association between blood glucose levels and AKI. However, future study may give more insight around the hypothesis we have generated. In conclusion, the onset of AKI in the CS setting still remains a relevant epidemiologic and clinical problem. In fact, the development of AKI is associated with a worse prognosis over time. We may assert that a predictive model, which encompasses pre-operative and intraoperative risk factors, provides a more comprehensive risk profile of patients at increased risk for AKI and subsequent mortality. Patients undergoing complex interventions, low blood glucose levels, acidosis, preexistent eGFR reduction, and long ECC and cross-clamp times have been found to have an increased risk for AKI. Further attention must be paid to younger patients too, who may present a more severe inflammatory response to surgical procedures. Moreover, patients with AKI, low blood glucose, ECC and cross-clamp duration are at increased mortality risk during the follow-up. Further investigations should clarify whether a better control of these risk factors would ultimately result in a protection against AKI and a better prognosis. Etiology of acute viral respiratory infections common in Pakistan: A review Summary Respiratory infections, especially those of the lower respiratory tract, remain a foremost cause of mortality and morbidity of children greater than 5 years in developing countries including Pakistan. Ignoring these acute‐level infections may lead to complications. Particularly in Pakistan, respiratory infections account for 20% to 30% of all deaths of children. Even though these infections are common, insufficiency of accessible data hinders development of a comprehensive summary of the problem. The purpose of this study was to determine the prevalence rate in various regions of Pakistan and also to recognize the existing viral strains responsible for viral respiratory infections through published data. Respiratory viruses are detected more frequently among rural dwellers in Pakistan. Lower tract infections are found to be more lethal. The associated pathogens comprise respiratory syncytial virus (RSV), human metapneumovirus (HMPV), coronavirus, enterovirus/rhinovirus, influenza virus, parainfluenza virus, adenovirus, and human bocavirus. RSV is more dominant and can be subtyped as RSV‐A and RSV‐B (BA‐9, BA‐10, and BA‐13). Influenza A (H1N1, H5N1, H3N2, and H1N1pdm09) and Influenza B are common among the Pakistani population. Generally, these strains are detected in a seasonal pattern with a high incidence during spring and winter time. The data presented include pneumonia, bronchiolitis, and influenza. This paper aims to emphasise the need for standard methods to record the incidence and etiology of associated pathogens in order to provide effective treatment against viral infections of the respiratory tract and to reduce death rates. pneumonia, bronchitis, bronchiolitis, and influenza. 1 Acute respiratory infections (ARI) are tremendous cause of health problems and mortality in emerging countries. 2 In South Asia, 48 of every 1000 children died before the age of five. 3 Pakistan is currently ranked as the sixth most populous country with the population of 199 million. 4 It is estimated that about 20% to 30% of all deaths of children under 5 years of age are because of respiratory infections in Pakistan. 5 lished between 2000 and 2018, using the key terms "viral respiratory infections," "influenza," "pneumonia," "bronchiolitis," "common cold," combined with the terms "morbidity," "mortality," "fatality," "epidemiology," "etiology," "genotypes," "developing countries," "South Asia," "Pakistan," "prevalence," "incidence" for articles published in English. Titles and abstracts of publications were reviewed, and articles containing data of viral respiratory infections from Pakistan were included. Studies were evaluated by all team members ( Figure 1). | Inclusion and exclusion criteria The inclusion criteria for studies were study of respiratory infections caused by viruses, data from Pakistan, publications from the year 2000 onwards in English language, and there was no restriction on subject age and gender ( Figure 1). The exclusion criteria were bacterial infections, information about countries other than Pakistan, and infections of non-human species. | RESULTS We initially identified 104 electronic articles from NCBI and Pakistan Journal of Chest Medicine that met our original search strategy. Figure 1 identifies the strategy used to screen articles for this review. After obtaining and evaluating each article, 12 articles were incorporated, and 92 were excluded. Out of 12 included articles, seven (including four for bronchiolitis and one for influenza) were about viral strains which were responsible for causing pneumonia in children and in older adults, and five were about surveillance and epidemiology of viral strains that cause influenza ( Table 1) | Included studies The characteristics of the articles included in this systematic review of the literature are shown in Table 1. | Etiology Viral etiology of LTRIs includes respiratory syncytial virus (RSV) and parainfluenza virus 3 among Pakistani population. 6 According to a study, enterovirus/rhinovirus, coronavirus, parainfluenza virus, adenovirus, and human Bocavirus were also found to be associated

with respiratory infections particularly with viral pneumonia. 7 Throat swabs were tested by RT-PCR. Nineteen percent of cases were identified as RSV associated infections. RSV appeared to be more incident from August to October and highest in September. Susceptibility increased in rainy seasons. 10 HMPV was detected in 7% of the cases in the above scenario, most of which were children under 1 year of age. 11 The results of a surveillance study conducted during January 2008 and December 2011 determined influenza A as a causative agent of infection in a great number of people. A total of 6258 samples were analyzed by RT-PCR assay, out of which 1489 samples were positive for influenza virus. Seventy-two percent of which were influenza A, and 28% were influenza B. Among influenza A, three strains were identified. These strains were A/H1N1, A/H3N2, and A/H1N1pdm09. 12 The results of another hospital-based study confirmed the viral association with ALRI. The study was conducted in two hospitals of twin cities of Pakistan from November 1986 to HMPV. RSV-A and RSV-B association was detected more commonly during 2 to 6 months of age, while influenza A was estimated as more common in 2.1 to 6-month age group. 9 RSV association with kids between the ages of one and 5 years old suffering from pneumonia and asthma was the most widely recognized finding of a 3-year study conducted in Karachi from August 2009 to June 2012. In older children between the ages of one and 5 years old, pneumonia and asthma were the most widely recognized findings. 10 In another study conducted by the group, the cases were characterized by pneumonia followed by bronchiolitis. 11 | Bronchiolitis Bronchiolitis is an acute viral infection, characterized by the inflammation of bronchioles. It is a contagious infection of the lower respiratory tract, and incidence is high during the winter season. It is more common in immuno-compromised individuals. 21 Chemical Stability Study of H1 Antihistaminic Drugs from the First and the Second Generations, Diphenhydramine, Azelastine and Bepotastine, in Pure APIs and in the Presence of Two Excipients, Citric Acid and Polyvinyl Alcohol The chemical stability of diphenhydramine (DIPH), azelastine (AZE) and bepotastine (BEPO) was examined in solutions and solids. The drugs were subjected to high temperature (70 °C for 35 h) or UV/VIS light (18.902–94.510 kJ/m2) at pH 1–13, to examine their percentage degradation and kinetics of degradation. Further, the stability of solid DIPH, AZE and BEPO was examined in the presence of excipients of different reactivity, i.e., citric acid (CA) and polyvinyl alcohol (PVA) under high temperature/high humidity (70 °C/80% RH) or UV/VIS light (94.510 kJ/m2). Under high temperature, DIPH degraded visibly (>19%) at pH 1 and 4, AZE was shown stable, while the degradation of BEPO was rather high (>17%) in all pH conditions. Under UV/VIS irradiation all the drugs were shown labile with degradation in the range 5.5–96.3%. As far as the solid mixtures were concerned, all drugs interacted with excipients, especially under high temperature/high humidity or UV/VIS light. As a result, DIPH, AZE and BEPO were compared in terms of their stability, with regard to their different structures and acid/base properties. All these results may be helpful for manufacturing, storing and applying these drugs in their topical (skin, nasal and ocular), oral and injectable formulations. Introduction H 1 antihistamines, formerly known as H 1 receptor antagonists or H 1 receptor blockers, are among the most commonly used medications for the prevention and treatment of rhinitis, conjunctivitis and urticaria, and other allergic and non-allergic diseases [1,2]. Moreover, quite new possible uses of these molecules are expected based on their antiviral, anticancer and other activities [3][4][5][6][7]. Thus, the H 1 antihistamines are considered of great actuality, because of their high efficacy in different therapeutic areas and present or future multiple uses. Diphenhydramine (DIPH) belongs to the first generation of antihistaminic drugs introduced in the 1940s, but more recently it has many applications because of its additional properties and despite the numerous side effects characteristic for this generation. DIPH is available at the market as oral liquids and tablets or topical formulations for allergic symptoms and itching. It is also widely used either orally or intravenously as an antiemetic drug and in cold preparations [2,8]. Azelastine (AZE), the drug from the second generation of antihistamines, is used for the treatment of allergic and vasomotor rhinitis, and allergic conjunctivitis. It is often administered topically, either as nasal sprays or ophthalmic solutions, and sometimes as oral tablets, e.g., in Japan. Its antihistaminic and mast cell-stabilizing effects are strengthened by its ability to inhibit inflammatory mediators including leukotrienes, kinins, cytokines and chemokines [2,9,10]. What is more, the antiviral activity of AZE was reported recently in SARS-CoV-2 infection, showing Although DIPH, AZE and BEPO have been present on the market for many years, there has been little research into their chemical stability, even for DIPH that has been used for over 50 years. Mainly, stability-indicating chromatographic methods have been published so far in which the forced degradation of DIPH was performed to confirm the selectivity of these methods [13][14][15][16][17]. The HPLC method is also recommended in the official monograph of DIPH in European Pharmacopoeia [18] for the separation of the related compounds of DIPH, i.e., Impurities A, B, C, D (benzhydrol) and E (benzophenone). In addition, the UHPLC assay of DIPH in the presence of its five related compounds, i.e., Impurities A, B, D, E and N-oxide of DIPH (DP1) has been elaborated [14] (Figure 2). Finally, two experiments including a stability study of DIPH in the presence of dextrose or NaCl in mini-bags and injection vials has been reported [15,16]. Although DIPH, AZE and BEPO have been present on the market for many years, there has been little research into their chemical stability, even for DIPH that has been used for over 50 years. Mainly, stability-indicating chromatographic methods have been published so far in which the forced degradation of DIPH was performed to confirm the selectivity of these methods [13][14][15][16][17]. The HPLC method is also recommended in the official monograph of DIPH in European Pharmacopoeia [18] for the separation of the related compounds of DIPH, i.e., Impurities A, B, C, D (benzhydrol) and E (benzophenone). In addition, the UHPLC assay of DIPH in the presence of its five related compounds, i.e., Impurities A, B, D, E and N-oxide of DIPH (DP1) has been elaborated [14] (Figure 2). Finally, two experiments including a stability study of DIPH in the presence of dextrose or NaCl in mini-bags and injection vials has been reported [15,16]. [14,18]. In the official monograph of AZE in European Pharmacopoeia, Impurities A, B, C, D and E are mentioned and the HPLC method is recommended for their separation [18]. As far as previous reports from the literature are concerned, one stability-indicating HPLC method for the determination of AZE and one HPLC method for separation of AZE from its degradation products has been reported. In Impurity D Impurity E DP1 Figure 2. Chemical structures of the related substances of diphenhydramine (DIPH) [14,18]. In the official monograph of AZE in European Pharmacopoeia, Impurities A, B, C, D and E are mentioned and the HPLC method is recommended for their separation [18]. As far as previous reports from the literature are concerned, one stability-indicating HPLC method for the determination of AZE and one HPLC method for separation of AZE from its degradation products has been reported. In addition, the structures of two degradants under acidic and alkaline (DP1), and oxidative (DP2) conditions have been proposed [19,20]. The HPLC method with gradient elution has also been reported for the separation of AZE and Impurities B, D and E [21]. In addition, spectrophotometric, TLC and HPLC methods have been developed for the determination of AZE in the presence of its alkaline degradant (DP3) [22]. The structures of the mentioned degradation products are shown in Figure 3. [14,18]. In the official monograph of AZE in European Pharmacopoeia, Impurities A, B, C, D and E are mentioned and the HPLC method is recommended for their separation [18]. As far as previous reports from the literature are concerned, one stability-indicating HPLC method for the determination of AZE and one HPLC method for separation of AZE from its degradation products has been reported. In addition, the structures of two degradants under acidic and alkaline (DP1), and oxidative (DP2) conditions have been proposed [19,20]. The HPLC method with gradient elution has also been reported for the separation of AZE and Impurities B, D and E [21]. In addition, spectrophotometric, TLC and HPLC methods have been developed for the determination of AZE in the presence of its alkaline degradant (DP3) [22]. The structures of the mentioned degradation products are shown in Figure 3. [18,20,22]. As far as BEPO is concerned, two reports were found in the literature presenting stability-indicating HPLC methods together with stress degradation experiments [23,24]. In addition, the sensitivity of BEPO to oxidation leading to one degradation product (DP1) has been confirmed using HPTLC, UHPLC and spectrophotometric methods [25]. When it comes to the photostability of DIPH, AZE and BEPO, the literary resources are even more limited. In a few studies [13,15,17,19,20,[22][23][24], these drugs were exposed to artificial UV light or natural sunlight, in solid state, solutions and liquid formulations, As far as BEPO is concerned, two reports were found in the literature presenting stability-indicating HPLC methods together with stress degradation experiments [23,24]. In addition, the sensitivity of BEPO to oxidation leading to one degradation product (DP1) has been confirmed using HPTLC, UHPLC and spectrophotometric methods [25]. When it comes to the photostability of DIPH, AZE and BEPO, the literary resources are even more limited. In a few studies [13,15,17,19,20,[22][23][24], these drugs were exposed to artificial UV light or natural sunlight, in solid state, solutions and liquid formulations, but the breadth of these experiments was rather scarce, and no kinetic aspects were taken into account. Only one paper from the literature presented the experiments on BEPO photostability and proposed the possible structures of its five degradation products (DP2-DP6) [26]. The possible structures of the degradants of BEPO which have been reported in the literature are presented in Figure 4. At the same time, it should be mentioned that the official monograph of BEPO has not been introduced to European Pharmacopoeia so far. The above data show that there is little information on the chemical stability and photostability of DIPH, AZE and BEPO under various stress conditions. Considering the importance of these drugs in therapy and the lack of sufficient reports concerning their chemical stability, the main goal of the present study was to examine their liability under different pH, temperature and light conditions. The first specific goal was to elucidate the degradation of DIPH, AZE and BEPO under high temperature or UV/VIS light in different media (0.1 M HCl, buffers of pH 4, 7, 10 and 0.1 M NaOH), together with quantitative measurements and kinetic investigations. The wide range of pH was used since the degradation processes of APIs can be highly dependent on their ionization. Moreover, pH-dependent degradation could be of great importance in relation to dissolution tests for respective oral formulations, as well as in relation to LADME processes, especially the liberation and absorption steps. Considering the numerous papers from the literature [13,20,22], we performed our experiments using a temperature of 70 • C. In addition, our preliminary studies showed that this temperature should give a sufficient degree of degradation to study the kinetics of the degradation of all tested drugs. but the breadth of these experiments was rather scarce, and no kinetic aspects were taken into account. Only one paper from the literature presented the experiments on BEPO photostability and proposed the possible structures of its five degradation products (DP2-DP6) [26]. The possible structures of the degradants of BEPO which have been reported in the literature are presented in Figure 4. At the same time, it should be mentioned that the official monograph of BEPO has not been introduced to European Pharmacopoeia so far. [25,26]. The above data show that there is little information on the chemical stability and photostability of DIPH, AZE and BEPO under various stress conditions. Considering the importance of these drugs in therapy and the lack of sufficient reports concerning their chemical stability, the main goal of the present study was to

examine their liability under different pH, temperature and light conditions. The first specific goal was to elucidate the degradation of DIPH, AZE and BEPO under high temperature or UV/VIS light in different media (0.1 M HCl, buffers of pH 4, 7, 10 and 0.1 M NaOH), together with quantitative measurements and kinetic investigations. The wide range of pH was used since the degradation processes of APIs can be highly dependent on their ionization. Moreover, pHdependent degradation could be of great importance in relation to dissolution tests for respective oral formulations, as well as in relation to LADME processes, especially the liberation and absorption steps. Considering the numerous papers from the literature [13,20,22], we performed our experiments using a temperature of 70 °C. In addition, our preliminary studies showed that this temperature should give a sufficient degree of degradation to study the kinetics of the degradation of all tested drugs. Active pharmaceutical ingredients (APIs) undergo chemical and physical changes when they are affected by external factors such as temperature, humidity, pH and light. Sensitivity of APIs to degradation varies with their chemical structures and reactivity, and the nature of the dosage forms in which they are formulated [27]. The last may be due to insufficient inertness of excipients that are used to obtain respective formulations. Generally, the excipients are inactive substances with regard to their biological activity for patients. However, they often contain chemical functional groups that can interact with APIs, sometimes leading to their degradation. In addition, they may contain or form their Active pharmaceutical ingredients (APIs) undergo chemical and physical changes when they are affected by external factors such as temperature, humidity, pH and light. Sensitivity of APIs to degradation varies with their chemical structures and reactivity, and the nature of the dosage forms in which they are formulated [27]. The last may be due to insufficient inertness of excipients that are used to obtain respective formulations. Generally, the excipients are inactive substances with regard to their biological activity for patients. However, they often contain chemical functional groups that can interact with APIs, sometimes leading to their degradation. In addition, they may contain or form their own specific degradants, and in turn, cause adverse chemical reactions. Thus, many guidelines emphasize the importance of testing the chemical stability of drugs as their bulk substances as their final pharmaceutical products, with particular respect to excipients [28,29]. It is obvious that experimental data the concerning stability of DIPH, AZE and BEPO in a solid state and their possible interactions with pharmaceutical excipients are too scarce. Thus, the next particular goal of the present study was to examine the chemical stability of DIPH, AZE and BEPO in the presence of two excipients of different reactivity, i.e., citric acid (CA) and polyvinyl alcohol (PVA). The solid mixtures of DIPH, AZE and BEPO with the mentioned excipients were prepared and then stressed with high temperature/high humidity or UV/VIS light, and finally analyzed using FT-IR and NIR spectroscopy. These excipients were chosen based on the literature data confirming their interactions with different APIs [30,31]. What is more, such combinations are present in real pharmaceutical products, for example ocular and nasal drops as well as solutions and tablets with DIPH, AZE and BEPO [32,33]. Moreover, they are often used in the manufacturing of new drug forms, especially as far as ocular and nasal drugs are concerned [34]. Their chemical groups potentially capable of interacting with other substances including APIs are shown in Figure 1. Results and Discussion Degradation and photodegradation of drugs are matters of great interest in modern pharmacy and therapy. A review of older APIs in these areas raises some concerns, as the data reported so far could be incomplete in the light of the new recommendations [35]. When it comes to the studies on DIPH, AZE and BEPO, the literary resources are clearly insufficient in these areas. Thus, we carried out a detailed investigation on the stability and photostability of DIPH, AZE and BEPO under various stress conditions. Microenvironment and its pH value can affect the chemical stability of drugs in bulk substances as well as in their dosage forms, and finally, their effectiveness and safety in patients. Thus, there is a great need to look deeply into the stability of APIs in different pH conditions to examine the risk of degradation or to detect new impurities. Moreover, the drugs may be stable to varying degrees in their ionized or non-ionized forms and may undergo specific acid-base catalyzed reactions [36]. Generally, the structures of the H 1 receptor antagonists present a diaryl substitution pattern and contain an amine function, both of which are essential for the H 1 receptor affinity. The amino moiety could be also important to obtain the corresponding salts of these basic drugs [1,2]. Chemically, DIPH belongs to antihistamines from the ethanolamine class, AZE is a phthalazine or methylazepane derivative, while for BEPO a metoxypiperidine moiety is essential ( Figure 1). As far as their acid-base properties are concerned, DIPH and AZE are both the weak bases with pKa of 8.88 and 8.98, respectively, while BEPO is a diprotic molecule with one acidic pKa equal 4.1 and one basic pKa equal 9.39 [37]. Methods for Quantitative Measurements For quantitative measurements, new reliable HPLC methods have been elaborated and validated according to the official guidelines [38,39]. These methods were found to be sufficiently selective, since they were able to separate DIPH, AZE and BEPO from their degradation products, and were sufficiently precise and accurate (Table 1). Meeting the official requirements made it possible to obtain reliable data for degradation kinetics of DIPH, AZE and BEPO. This, in turn, is extremely important in terms of predicting the shelf life of the drugs and their safety for patients [40]. Thermal Degradation in Solutions Our study showed that the maximum of thermal degradation of DIPH occurred in 0.1 M HCl (>30%) and pH 4 (>19%), where its amino group was highly protonized. In other media, i.e., buffers of pH 7, 10 and 0.1 M NaOH, the drug seemed to be more stable in its non-ionized form ( Table 2). As was noticed above, some reports on the stability of DIPH have been published so far, but they are not consistent and comprehensive. DIPH was shown to be stable in the study of Al-Salman et al. [13] with degradation below 5%, when exposed to acidic, alkaline and thermolytic conditions for 2-3 h. In the study of Sabins et al. [16] degradation of DIPH was shown below 20%, both in strongly acidic and strongly alkaline media. On the contrary, in the study of Donelly [15] DIPH decomposed more in acidic than alkaline conditions. This was also shown in the present study (Table 2). AZE presented similar basic properties to DIPH with a similar pKa value 8.98. However, its sensitivity to thermal stress was to some extent different. Our study confirmed its higher stability in a strong acidic medium, which is consistent with previously reported results [19]. What is more, AZE was stable in a strong alkaline medium (Table 2). However, in the study of El-Shaheny at al. [20], the visible degradation of AZE was observed in both acidic and alkaline conditions. A diprotic BEPO was shown to be more sensitive to degradation then DIPH and AZE in a whole pH range with 17.4-56.8% of degradation. It was extremely labile in 0.1 M HCl and 0.1 M NaOH in both its ionized forms ( Table 2). According to the literature, acid and base stressing of BEPO at an ambient temperature led to a degradation of 73.7% in 0.1 M HCl and in 17% in 5 M NaOH [24]. Table 2. Parameters of HPLC-UV methods for the quantitative determination of diphenhydramine (DIPH), azelastine (AZE) and bepotastine (BEPO). Conditions Degradation Photolytic Degradation in Solutions UV/VIS spectral analysis that is recommended by the official guidelines as the first step of testing the drug photoreactivity and phototoxicity [41] was performed for DIPH, AZE and BEPO. DIPH was shown as absorbing appreciably over the range 200-240 nm while above 240 nm the absorbance decreased gradually to 270 nm. The AZE spectrum was characterized by bands over the range 200-340 nm with three peaks at 215, 258 and 293 nm. BEPO was shown as absorbing appreciably over the range 200-270 nm while above 270 nm the absorbance decreased gradually to 370 nm ( Figure 5A). Thus, at least AZE and BEPO showed bands that overlapped with the light spectrum used in the present study, i.e., 300-800 nm. Our study showed the sensitivity of DIPH to UV/VIS light, independently of pH value, in its ionized and non-ionized forms. Our experiments led to the degradation of DIPH above 40% with the maximum of decomposition in 0.1 M HCl (56.5%) and a buffer of pH 10 (49.5%). (Table 3). Previously, DIPH was shown to be stable in the standard tests confirming photostability, under energy of 200 Wh/m 2 and 1,200,000 lux h [13,15]. On the contrary, gradual decomposition was observed under forced photolytic conditions in the study of Bober [17] similar to our results. light at wavelengths of about 300 nm and above ( Figure 5A). In addition, some visible changes were observed in the UV spectrum of AZE after irradiation in solutions of pH 10-13, where the maximum of photodegradation of AZE (>90%) was observed ( Figure 5B). These data were in agreement with previously reported results from the literature confirming the photosensitivity of AZE in methanolic solution [19]. At the same time, it was interesting to observe that AZE was shown to be more sensitive to UV/VIS irradiation than to high temperatures at similar pH conditions. Unlike the case of thermolytic conditions, our study showed the lability of AZE in photolytic conditions (Table 3), which can be explained by the ability of AZE to absorb light at wavelengths of about 300 nm and above ( Figure 5A). In addition, some visible changes were observed in the UV spectrum of AZE after irradiation in solutions of pH 10-13, where the maximum of photodegradation of AZE (>90%) was observed ( Figure 5B). These data were in agreement with previously reported results from the literature confirming the photosensitivity of AZE in methanolic solution [19]. At the same time, it was interesting to observe that AZE was shown to be more sensitive to UV/VIS irradiation than to high temperatures at similar pH conditions. A diprotic BEPO was shown to be more sensitive to photodegradation then DIPH and AZE (Table 3). All pH conditions used in our experiments led to the degradation of BEPO above 20% with the maximum of decomposition in 0.1 M HCl (96.3%) and 0.1 M NaOH (61.7%). Thus, the sensitivity of BEPO to UV/VIS light was documented for its protonated as well as for anionic forms. Previously, BEPO was shown to be stable in methanolic solution under standard ICH conditions of 200 Wh/m 2 and 1,200,000 lux h [23,24]. However, irradiation of BEPO with 0.8 W/m 2 and 4500 lux h and under basic conditions led to degradation of 23.9% [26]. Bearing in mind all above results, we can conclude that all pH conditions used in our experiments led to photodegradation of DIPH, AZE and BEPO. It is worth mentioning that the pH range 4-10 covers, at least in part, the pH range suitable for ophthalmic drugs. Thus, it may increase the risk of photodegradation of DIPH, AZE and BEPO during the manufacture and useof respective ocular solutions, and in consequence, decrease their therapeutic efficacy in patients. Kinetics Kinetic parameters were calculated when the percentage degradation of DIPH, AZE and BEPO was at least 10% in the assumed time of experiment. The plots of logarithms of concentration of non-degraded drugs versus time of degradation showed stronger correlations (higher r 2 values) than the plots of concentrations of non-degraded drugs versus time of degradation, confirming the pseudo-first-order kinetics of these degradation processes with the rate constants in the range 10 −4 -10 −3 min −1 . The calculated t 0.5 values for thermal degradation of DIPH varied from 18.01 h (at pH 1), through 50.15 h (at pH 4) to 71.65 h (at pH 7), confirming considerable degradation of DIPH in acidic conditions in its ionized form ( Table 2). Respective values of t 0.5 for photodegradation processes were in the range 12.39-16.72 h, over

the pH range 1-13. The calculated t 0.5 values for AZE photodecomposition varied from 50.15 h (at pH 4), through 12.53 h (at pH 7) to 4.18 h (at pH 10-13), confirming the considerable degradation of AZE in alkaline conditions where the drug was electrically neutral. At the same time, the first order rate constants obtained at pH 10-13 were one order of magnitude higher (10 −3 min −1 ) than those obtained at lower pH values ( Table 3) Table 3), confirming that BEPO was prone to degradation in both its ionized forms. At the same time, the first order rate constant obtained in photolytic stress at pH 1 was one order of magnitude higher (10 −3 min −1 ) than that obtained at other pH values. The influence of pH on the reaction rate constants during thermolytic or photolytic degradation for DIPH, AZE and BEPO is depicted in Figure 6. It is interesting to observe that the shape of pH-rate profile of BEPO was more complex than that of DIPH and AZE, with the maxima in the extreme pH after both thermolytic and photolytic stress. Table 3), confirming that BEPO was prone to degradation in both its ionized forms. At the same time, the first order rate constant obtained in photolytic stress at pH 1 was one order of magnitude higher (10 −3 min −1 ) than that obtained at other pH values. The influence of pH on the reaction rate constants during thermolytic or photolytic degradation for DIPH, AZE and BEPO is depicted in Figure 6. It is interesting to observe that the shape of pHrate profile of BEPO was more complex than that of DIPH and AZE, with the maxima in the extreme pH after both thermolytic and photolytic stress. HPLC-UV Chromatograms When HPLC-UV chromatograms of the stressed solutions of DIPH were examined, at least four degradation products, i.e., D1 and D4 with the retention times 0.935 and 7.120 min (in thermolytic conditions) and D2 and D3 with the retention times 1.893 and 4.509 min (in both thermolytic and photolytic conditions) were noticed. It was interesting to observe that the D4 product was detected only in the samples at pH ≤ 7, while the other products, i.e., D1-D3 were detected in most of the stressed samples. A representative HPLC-UV Chromatograms When HPLC-UV chromatograms of the stressed solutions of DIPH were examined, at least four degradation products, i.e., D1 and D4 with the retention times 0.935 and 7.120 min (in thermolytic conditions) and D2 and D3 with the retention times 1.893 and 4.509 min (in both thermolytic and photolytic conditions) were noticed. It was interesting to observe that the D4 product was detected only in the samples at pH ≤ 7, while the other products, i.e., D1-D3 were detected in most of the stressed samples. A representative chromatogram showing the disappearance of the peak corresponding to unmodified DIPH and additional peaks from the decomposition products D1-D4 is presented in Figure 7A. HPLC-UV Chromatograms When HPLC-UV chromatograms of the stressed solutions of DIPH were examined, at least four degradation products, i.e., D1 and D4 with the retention times 0.935 and 7.120 min (in thermolytic conditions) and D2 and D3 with the retention times 1.893 and 4.509 min (in both thermolytic and photolytic conditions) were noticed. It was interesting to observe that the D4 product was detected only in the samples at pH ≤ 7, while the other products, i.e., D1-D3 were detected in most of the stressed samples. A representative chromatogram showing the disappearance of the peak corresponding to unmodified DIPH and additional peaks from the decomposition products D1-D4 is presented in Figure 7A. In the study from the literature, AZE was exposed to acidic and alkaline conditions, and produced the same degradation product DP1 (Figure 3) [20]. In our study, thermal stress did not lead to the detection of the peaks of degradants on respective chromatograms. However, the photodecomposition of AZE led to at least four degradation products being obtained. In a whole pH range 1-13, A1 and A4 products were observed with respective retention times of 1.036 and 2.769 min while non-degraded AZE showed the retention time 3.054 min. At pH 10-13 where the decomposition of the drug was higher, two next products were detected at their retention times 1.542 min (A2) and 2.131 min (A3) ( Figure 7B). When chromatograms of the stressed solutions of BEPO were examined, it was observed that it decomposed forming at least two degradation products, B1 with the retention time 2.415 min and B2 with the retention time 2.793 min. These two products were detected after stressing BEPO at extreme pH values, as under thermolytic as photolytic stress. A representative chromatogram showing the disappearance of the peak corresponding to unmodified BEPO and two additional peaks from the decomposition products B1 and B2 is presented in Figure 7C. One study from the literature allowed the detection of two degradants of BEPO after storing in acidic or oxidative conditions [24,25]. In the next studies from the literature [26], irradiation under alkaline conditions led to the formation of five degradants (Figure 4). Thus, our results about BEPO degradation at extreme pH values were confirmed. Taken the above information together, it was concluded that DIPH and BEPO showed the possibility of both thermal and photodecomposition. At the same time, AZE was shown to be sensitive to UV/VIS light, although it was more stable under thermolytic conditions. According to the literature, the decomposition of DIPH can be related, mainly in part, to the cleavage of the ether linkage under acidic conditions. Oxidation of DIPH is also possible where the amine group is the most likely site of oxidation with generation the corresponding N-oxide [15,18]. Because there is very little information on the photolytic decomposition of DIPH [42], the question is whether such mechanisms also occur during photolytic stress. Taking into account the obtained results and the available data from the literature, we propose a probable decomposition of DIPH through the breakdown of the ether bond leading to diphenylmethanol (benzhydrol), diphenylmethanone (benzophenone) or diphenylmethane. As a probable degradation method, we can also indicate demethylation leading to the formation of the product with the secondary amino group (Figure 8). also possible where the amine group is the most likely site of oxidation with generat the corresponding N-oxide [15,18]. Because there is very little information on the pho lytic decomposition of DIPH [42], the question is whether such mechanisms also oc during photolytic stress. Taking into account the obtained results and the available d from the literature, we propose a probable decomposition of DIPH through the bre down of the ether bond leading to diphenylmethanol (benzhydrol), diphenylmethan (benzophenone) or diphenylmethane. As a probable degradation method, we can also dicate demethylation leading to the formation of the product with the secondary am group (Figure 8). As far as AZE is concerned, acidic degradation through the opening of azepane ring and oxidative demethylation or alkaline degradation leading to opening of phthalazine moiety has been proposed in the literature [18,20,22]. Based on the obtained results, we can propose the photodegradation of AZE by hydrolysis of the amide bond in the phthalazine ring and also by demethylation in the azepane moiety ( Figure 9). Molecules 2022, 27, x FOR PEER REVIEW As far as AZE is concerned, acidic degradation through the opening of a and oxidative demethylation or alkaline degradation leading to opening of moiety has been proposed in the literature [18,20,22]. Based on the obtained can propose the photodegradation of AZE by hydrolysis of the amide bond in azine ring and also by demethylation in the azepane moiety ( Figure 9). According to the literature, oxidative and acidic degradation of BEPO c to the oxidation or cleavage of butanoic acid from the structure [24,25]. Photo of BEPO was proposed by the cleavage of its ether bond, especially in stron tions. Moreover, the plausibility of the addition of an oxygen atom to form Nof the nitrogen atoms was suggested [26]. Based on the obtained results, we BEPO decomposition through the disintegration of the ether bond with the f the appropriate products ( Figure 10). According to the literature, oxidative and acidic degradation of BEPO could be due to the oxidation or cleavage of butanoic acid from the structure [24,25]. Photodegradation of BEPO was proposed by the cleavage of its ether bond, especially in strong pH conditions. Moreover, the plausibility of the addition of an oxygen atom to form N-oxide at any of the nitrogen atoms was suggested [26]. Based on the obtained results, we can propose BEPO decomposition through the disintegration of the ether bond with the formation of the appropriate products ( Figure 10). According to the literature, oxidative and acidic degradation of BEPO could to the oxidation or cleavage of butanoic acid from the structure [24,25]. Photodegr of BEPO was proposed by the cleavage of its ether bond, especially in strong pH tions. Moreover, the plausibility of the addition of an oxygen atom to form N-oxid of the nitrogen atoms was suggested [26]. Based on the obtained results, we can p BEPO decomposition through the disintegration of the ether bond with the form the appropriate products ( Figure 10). At the same time, we realize that further experiments allowing identification and more detailed studies on degradants of DIPH, AZE and BEPO should be performed. This is especially important for potential photodecomposition products. It is well documented that their phototoxic and photoallergic effects can occur in patients, especially in tissues exposed to sunlight [43,44]. DIPH, AZE and BEPO, which can be used in the eye drops, external solutions and ointments may degrade upon extensive light exposure with the lowering of their therapeutic action or even generating toxic products. What is more, the stability of ophthalmic solutions is the most important due to eye sensitivity [45]. These issues give also rise to a need for optimal pH selection for DIPH, AZE and BEPO in their formulations. Degradation in Solids and Impact of Excipients Official guidelines emphasize the importance of testing chemical stability of drugs not only in their bulk substances but also in their final pharmaceutical preparations, in the presence of excipients. These excipients should be inert by definition, but in many cases, they may react or influence reactions taking place in the microenvironment of the preparations [29]. For this reason, two component solid mixtures of DIPH, AZE and BEPO with two excipients of different reactivity, i.e., citric acid (CA) and polyvinyl alcohol (PVA), were prepared to examine the possible impacts of these excipients on APIs stability. High temperature and high humidity (70 • C and 80% RH for 35 h) but also UV/VIS light (94.510 kJ/m 2 ) were applied as stressors in order to accelerate these potential interactions and finally, FT-IR and NIR spectroscopy were used to examine the samples. At the beginning, differences between the stressed and non-stressed bulk substances of DIPH, AZE, BEPO, CA and PVA were taken into account based on their FT-IR characteristics presented in Table 4 and the NIR characteristics from the literature [46]. In the next step, the differences between the stressed and non-stressed physical mixtures of each API with each excipient were examined. Below we present the results for these mixtures where there were visible changes suggesting interactions in a respective drug-excipient pair. FT-IR and NIR spectra that showed visible changes due to potential interactions are presented in Figures 11-15. As far as previous studies from the literature are concerned, DIPH in powder was stored in a hot air at 105 • C for 2-3 h in the study of Al-Salman et al. [13]. Solid AZE was exposed to UV radiation at 254 nm for 20 h in the study of El-Shaheny et al. [20]. BEPO in powder was exposed to thermal stress at 60 • C for 21 days and radiation of 1000 W/m 2 and 6,000,000 lux h in the study of Singh [26]. All these authors concluded no degradation of DIPH, AZE and BEPO in a solid state, although mainly chromatographic methods were used for these previous experiments. Our study using FT-IR and NIR spectroscopy confirmed the results about the stability of DIPH, AZE and BEPO as bulk substances; however, some changes were observed in the presence of the excipients, i.e., CA and PVA. of DIPH and CA

was treated with high temperature and high humidity, the new changes concerning the peaks at 3034, 3016 and 2951 cm −1 were observed as lowering and shifting to higher wavenumbers. Moreover, the peaks due to C-C stretching vibrations in aromatic rings at 1468 and 1455 cm −1 changed their shapes. What is more, the sharp peak at 1171 cm −1 due to the C-O-C stretching vibration in the ether group of DIPH was much lower. At the same time, a broad band of CA between 3500 and 3200 cm −1 became more dominant ( Figure 11C). Based on these changes, we can suppose that the ether group of DIPH could be affected in the presence of CA. As was stated above, the instability of DIPH can be related to cleavage of the ether linkage under acidic conditions (Figure 8). In addition, CA can interact with basic drugs and form hydrogen bonds. Such changes were described previously for haloperidol and itraconazole in their solid dispersions with CA [47,48]. As it was shown in Figure 8, DIPH can also degrade to the secondary amine. Thus, the possibility of hydrogen bonding between CA and such decomposed DIPH could be suggested. This groups in the range 3500-3200 cm −1 . Unfortunately, these overlapped CA bands made it difficult to confirm whether secondary amine bands appeared as a consequence of degradation of DIPH ( Figure 11C). Our research also showed some changes at the NIR spectra of DIPH after irradiating the drug in presence of PVA, i.e., changing the shape of the band at 6275 cm −1 (the first overtones of C-H stretching vibrations), lowering the band at 5161 cm −1 (the second overtones of C-O stretching vibrations) and disappearing the band at 4332 cm −1 (Figure 12A In order to obtain more reliable results, the NIR spectra of DIPH and PVA alone, before and after irradiation, were checked carefully. Because the changes were not observed for single substances, some interactions between DIPH and PVA that were initiated or accelerated by UV/VIS light could be taken into account. As far as AZE in the mixtures with CA and PVA was concerned, the observed changes were less numerous as compared to DIPH. After mixing AZE with CA at ambient conditions, most of the characteristic bands and peaks of the drug were still clearly seen. When the mixture of AZE and PVP was examined without any stress, the peak of AZE at ca. 3400 cm −1 was overlapped with the signals of OH groups of PVA at 3600-3300 cm −1 (Figure 14A,B). When UV/VIS light was used for this mixture, a new band at 1727 in the spectrum of the non-stressed mixture ( Figure 14C). Thus, some interactions of AZE with PVA were possible, especially under UV/VIS irradiation, but their intensities were not sufficient to be interpreted in the presented conditions. In the study from the literature, the photodecomposition of the amide bond in the AZE structure was suggested, similar to the thermal decomposition [20]. However, the spectra obtained for the mixture with PVA did not show changes at wavenumbers corresponding to the C=O amide group vibrations of AZE. group) vibrations at 1716 cm −1 was overlapped with the peak of PVA at 1725 cm −1 , while other peaks of BEPO, e.g., those due to C-O-C stretching vibrations at 1290 and 903 cm −1 were significantly lowered ( Figure 15A,B). When the irradiation of BEPO in the presence of PVA was performed, additive changes were observed as disappearing the bands of BEPO at 3416 and 3349 cm −1 . What is more, the overlapped peak at ca. 1720 cm −1 changed its shape while the peak of BEPO due to C=O stretching (carboxylic acids) vibrations at 1716 cm −1 became visible as a characteristic sharp peak again ( Figure 15C). Thus, some interactions including the carboxylic group and the ether group of BEPO and the OH groups of PVA could be possible. As was mentioned above, the ether group of BEPO could be affected under photolytic conditions ( Figure 10). It could be supposed that the presence of PVP could facilitate such changes. Bearing in mind all above results, it could be concluded that both BEPO and DIPH containing the ether groups seemed to be less stable than AZE and interacted much more easily with the excipients. Another interesting conclusion was that the impact of CA on stability of DIPH, AZE and BEPO was higher under thermolytic stress while the impact of PVA was higher under photolytic stress. Bearing in mind UV absorptive properties of PVA above 300 nm, it could be speculated that PVA can produce its own photodegradants capable of accelerating the degra- After mixing DIPH and CA, the most characteristic bands of the drug including these due to the signals of C-O-C and C-N (N-CH 3 ) groups at 1171 and 1104 cm −1 were still clearly seen. However, some other bands changed due to overlapping with some bands of CA. This was seen for the peaks of DIPH at 3034, 3016 and 2951 cm −1 that overlapped with OH stretching vibrations from carboxylic group of CA at 3050 cm −1 . Some other bands of DIPH were lacking, e.g., those due to C-H stretching (aromatic rings) at 2592 cm −1 and the neighboring peaks at 2517, 2483 and 2452 cm −1 (Figure 11A,B). When the mixture of DIPH and CA was treated with high temperature and high humidity, the new changes concerning the peaks at 3034, 3016 and 2951 cm −1 were observed as lowering and shifting to higher wavenumbers. Moreover, the peaks due to C-C stretching vibrations in aromatic rings at 1468 and 1455 cm −1 changed their shapes. What is more, the sharp peak at 1171 cm −1 due to the C-O-C stretching vibration in the ether group of DIPH was much lower. At the same time, a broad band of CA between 3500 and 3200 cm −1 became more dominant ( Figure 11C). Based on these changes, we can suppose that the ether group of DIPH could be affected in the presence of CA. As was stated above, the instability of DIPH can be related to cleavage of the ether linkage under acidic conditions (Figure 8). In addition, CA can interact with basic drugs and form hydrogen bonds. Such changes were described previously for haloperidol and itraconazole in their solid dispersions with CA [47,48]. As it was shown in Figure 8, DIPH can also degrade to the secondary amine. Thus, the possibility of hydrogen bonding between CA and such decomposed DIPH could be suggested. This can also be proposed from the changes observed within the CA bands due to OH (alcohol) groups in the range 3500-3200 cm −1 . Unfortunately, these overlapped CA bands made it difficult to confirm whether secondary amine bands appeared as a consequence of degradation of DIPH ( Figure 11C). Our research also showed some changes at the NIR spectra of DIPH after irradiating the drug in presence of PVA, i.e., changing the shape of the band at 6275 cm −1 (the first overtones of C-H stretching vibrations), lowering the band at 5161 cm −1 (the second overtones of C-O stretching vibrations) and disappearing the band at 4332 cm −1 (Figure 12A-C). In order to obtain more reliable results, the NIR spectra of DIPH and PVA alone, before and after irradiation, were checked carefully. Because the changes were not observed for single substances, some interactions between DIPH and PVA that were initiated or accelerated by UV/VIS light could be taken into account. As far as AZE in the mixtures with CA and PVA was concerned, the observed changes were less numerous as compared to DIPH. After mixing AZE with CA at ambient conditions, most of the characteristic bands and peaks of the drug were still clearly seen. However, the peak of AZE at 3400 cm −1 was partly overlapped with the signal of OH (alcohol) group of CA in the range 3275-3380 cm −1 ( Figure 13A,B). In the mixture of AZE and CA subjected to thermal and moisture degradation, the bands between 3400 and 3380 cm −1 changed their shapes even more visibly. In addition, sharp peaks in the range 1650 and 1550 cm −1 were fused in one broad peak ( Figure 13C). Although the overlapping of C=O amide group vibrations of AZE and of C=O carboxylic group vibrations of CA made it difficult to interpret the nature of these changes, it could be supposed that the amide group of AZE was affected in acidic conditions (Figure 9). The possibility of the acidic decomposition of the amide group of AZE has also been proposed in the literature [20]. When the mixture of AZE and PVP was examined without any stress, the peak of AZE at ca. 3400 cm −1 was overlapped with the signals of OH groups of PVA at 3600-3300 cm −1 ( Figure 14A,B). When UV/VIS light was used for this mixture, a new band at 1727 cm −1 occurred, probably due to the C-O stretching vibrations of PVA which was not seen in the spectrum of the non-stressed mixture ( Figure 14C). Thus, some interactions of AZE with PVA were possible, especially under UV/VIS irradiation, but their intensities were not sufficient to be interpreted in the presented conditions. In the study from the literature, the photodecomposition of the amide bond in the AZE structure was suggested, similar to the thermal decomposition [20]. However, the spectra obtained for the mixture with PVA did not show changes at wavenumbers corresponding to the C=O amide group vibrations of AZE. When the mixture of BEPO and PVP was examined without any stress, the peaks of BEPO at 3416 and 3349 cm −1 were overlapped with the signal of OH group of PVA at 3275 cm −1 . What is more, the overlapped peak of BEPO due to C=O stretching (carboxylic group) vibrations at 1716 cm −1 was overlapped with the peak of PVA at 1725 cm −1 , while other peaks of BEPO, e.g., those due to C-O-C stretching vibrations at 1290 and 903 cm −1 were significantly lowered ( Figure 15A,B). When the irradiation of BEPO in the presence of PVA was performed, additive changes were observed as disappearing the bands of BEPO at 3416 and 3349 cm −1 . What is more, the overlapped peak at ca. 1720 cm −1 changed its shape while the peak of BEPO due to C=O stretching (carboxylic acids) vibrations at 1716 cm −1 became visible as a characteristic sharp peak again ( Figure 15C). Thus, some interactions including the carboxylic group and the ether group of BEPO and the OH groups of PVA could be possible. As was mentioned above, the ether group of BEPO could be affected under photolytic conditions ( Figure 10). It could be supposed that the presence of PVP could facilitate such changes. Bearing in mind all above results, it could be concluded that both BEPO and DIPH containing the ether groups seemed to be less stable than AZE and interacted much more easily with the excipients. Another interesting conclusion was that the impact of CA on stability of DIPH, AZE and BEPO was higher under thermolytic stress while the impact of PVA was higher under photolytic stress. Bearing in mind UV absorptive properties of PVA above 300 nm, it could be speculated that PVA can produce its own photodegradants capable of accelerating the degradation of sensitive APIs. This is even more important because such combinations are present in real pharmaceutical products with DIPH, AZE or BEPO [32,33]. Moreover, new formulations containing DIPH, AZE or BEPO are being designed or submitted for registration including oral pharmaceutical compositions as well as nasal and ocular formulations which contain CA or PVA [49]. Thus, a suggestion that the utilization of respective excipients requires careful evaluation on a case-by-case basis could be proposed. Materials Pharmaceutical grade standards of diphenhydramine hydrochloride (DIPH), azelastine hydrochloride (AZE) and bepotastine besylate (BEPO), papaverine and xylomethazoline hydrochlorides (the internal standards for HPLC methods) from Sigma-Aldrich (St. Louis, MO, USA), acetonitrile, methanol and water for chromatography from Merck (Darmstad, Germany), glacial acetic acid, sodium acetate, hydrochloric acid, sodium chloride, sodium tetraborate, phosphoric acid, sodium hydrogen phosphate, sodium dihydrogen phosphate, kalium dihydrogen phosphate and kalium hydroxide from POCh (Gliwice, Poland) were

used. The buffer solutions, i.e., acetate buffer of pH 4, phosphate buffer of 7 and borate buffer of pH 10, were used as degradation media. Phosphate buffer of pH 3 was used for preparing the mobile phases in our HPLC methods. All buffers were prepared according to European Pharmacopoeia [18]. Apparatus for Accelerated Degradation A climate chamber KBF-LQC (Binder GmbH, Tuttlingen, Germany) set at 70 • C and 80% RH was used for high temperature and high humidity stress. A Suntest CPS Plus chamber from Atlas (Linsengericht, Germany) was used as a solar simulator of UV/VIS light in the range 300-800 nm. This chamber was equipped with an appropriate temperature control unit and the temperature was not higher than 35 • C during all experiments. Degradation of DIPH, AZE and BEPO in Solutions Volumes of 1 mL from the stock solutions of DIPH, AZE or BEPO (1 mg/mL) were dispensed in duplicate to standardized quartz glass dishes and mixed with 1 mL of respective medium (0.1 M HCl, 0.1 M NaOH, buffers of pH 4, 7 and 10). The stressed conditions were established based on respective ICH recommendations [50]. The samples were exposed to a high temperature and high humidity (70 • C and 80% RH) for [51]. The controls (non-stressed samples) were stored at ambient conditions (23 ± 2 • C) in a dark place. After finishing forced degradation, the samples of DIPH, AZE and BEPO were diluted with methanol to cover the linearity ranges of our HPLC methods and analyzed quantitatively as was described below. The quantitative assays were repeated three times for each sample. The concentrations of non-degraded drugs were calculated from respective calibration equations. Degradation of DIPH, AZE and BEPO in Solids Binary mixtures of DIPH, AZE and BEPO with two excipients, i.e., CA and PVA were prepared by mixing the components in an agate mortar at 1:1 ratio (w/w). Then, drugs alone, excipients alone and the prepared mixtures were dispersed in duplicate as ca. 20 mg portions to the standardized quartz glass flat vessels. Half of them were placed in a climate chamber set at 70 • C and 80% RH for 35 h while the rest of the samples were placed in a solar simulator and irradiated with the energy equal 94.510 kJ/m 2 . The controls (nonstressed samples) were stored in a dessicator at ambient conditions (23 ± 2 • C) in a dark place. After finishing accelerated degradation, the samples were analyzed qualitatively using the FT-IR and NIR methods as was described below. Spectrophotometric Measurements DIPH, AZE and BEPO as bulk substances were dissolved in methanol, diluted to a final concentration of 10 µg/mL and analyzed spectrophotometrically in the range 200-400 nm to obtain the UV absorption spectra and estimate the risk of their photodegradation. A UV-2001 double beam spectrophotometer from Hitachi High-Technologies Corporation (Tokyo, Japan) and quartz cuvettes with a 1 cm path length were used for these measurements. Kinetics of Degradation in Solutions In order to obtain the data for kinetic calculations the stressed liquid samples of DIPH, AZE and BEPO were analyzed quantitatively using HPLC methods described below. The concentrations of the drugs remaining after each period of degradation or logarithms of these concentrations were plotted against corresponding time of degradation to obtain the equations y = ax + b, the determination coefficients r 2 and the reaction order. Finally, further kinetic parameters, i.e., degradation rate constant (k) and degradation time of 50% substance (t 0.5 ) were calculated using respective formula. Chromatography Analyses were performed with a model 515 pump, a Rheodyne 20 µL injector and a model UV 2487 DAD, controlled by Empower 3 software, all from Waters UK Sales (Elstree, UK). Separation was carried out on a LiChrospher ® RP-8 column (125 × 4.0 mm, 5 µm) from Merck. The mobile phase for DIPH was the mixture of acetonitrile and phosphate buffer of pH 3 (50:50, v/v) while for AZE and BEPO, the mobile phase consisted of acetonitrile, methanol and phosphate buffer of pH 3 (50:30:20, v/v/v). The flow rate of the mobile phases was always 2 mL/min and the UV detection was set at 220 nm for all tested drugs. For DIPH and AZE determination, papaverine was used as internal standard while for BEPO analysis, xylomethazoline was a better option. Validation of the Methods Selectivity of the methods was examined by the determination of DIPH, AZE and BEPO in the presence of their degradation products in the stressed samples. For calibration, working solutions of the drugs were prepared by dispensing 0.1-1.0 mL volumes from the stock solutions (1 mg/mL) to 10 mL volumetric flasks to reach the concentration range 10-100 µg/mL. Internal standards (i.s.) were added in 0.2-0.5 mL volumes of the corresponding stock solutions (1 mg/mL). After adjusting with methanol to the marks, samples were injected onto the column in 6 repetitions. The ratio of peak areas (the drug and respective i.s.) were plotted against the corresponding concentration of the drug to construct the calibration equations. The limit of detection (LOD) and the limit of quantification (LOQ) were determined from the SD of the intercept and the slope of the respective regression lines. In order to verify accuracy and repeatability, replicate injections of working solutions of the drugs at low (15 µg/mL), medium (50 µg/mL) and high (90 µg/mL) concentrations were conducted. The accuracy was calculated as the percentage of the analytes recovered by respective assay while repeatability was considered as the relative standard deviation for the repeated intra-day and inter-day estimations. FT-IR and NIR Measurements FT-IR and NIR spectra were recorded on a Nicolet 6700 spectrometer (ThermoScientific, Waltham, MA, USA), equipped with a Smart iTR™ ATR Sampling Accessory and Near IR Integrating Sphere, respectively. After recording a background spectrum, the samples were analyzed over the ranges 4000-650 cm −1 for FT-IR and 10,000-4000 cm −1 for NIR spectroscopy. Each spectrum was recorded as an average of four scans and analyzed using OMNIC software from ThermoScientific. Conclusions As was described above, there are not many papers in the literature concerning chemical stability with respect to DIPH, AZE and BEPO. Thus, the results presented here supplemented the literature resources in these areas. As a result, we reported the comprehensive stability data for DIPH, AZE and BEPO in solutions at different pH together with their kinetics of degradation as well as their stability as solids in the presence of two different excipients. In solutions, AZE was shown as the most stable compound as far as the impact of temperature was concerned while DIPH and BEPO containing the ether groups in their structures were shown to be more labile. However, DIPH, AZE and BEPO were all shown to be sensitive to UV/VIS light with the percentage degradation depending on the pH of the environment. In solids, DIPH, AZE and BEPO were shown to be stabile but they showed some interactions with CA and PVP that were intensified in thermolytic and photolytic conditions. Possible decomposition of DIPH and BEPO could happen due to the ether groups breaking. As far as AZE is concerned, its amide group can undergo hydrolysis, especially under photolytic conditions. N-demethylation could be also proposed for DIPH and AZE. In addition, interesting results were obtained indicating that CA can affect the stability of the drugs under thermolytic conditions, while PVA had a stronger impact under UV/VIS irradiation. Many new pharmaceutical formulations with H 1 antihistamines are being introduced to the market, depending on the route of the drug administration, in tablets, oral solutions, topical solutions, ointments, creams and very often in ocular drops. Thus, the results presented here can be utilized to improve the quality of respective formulations of DIPH, AZE and BEPO by minimizing the risk associated with their degradation. The dependence of the chemical stability of DIPH, AZE and BEPO on their structures and the presence of the respective functional groups together with differences in their acid-base properties were taken into account. This may be an inspiration in the design of new drugs from the H 1 antihistamines because the most of the newest H 1 antihistamines are introduced as structural modifications of preexisting drug substances. High-dose-rate brachytherapy delivered in two fractions as monotherapy for low-risk prostate cancer Purpose High-dose-rate (HDR) brachytherapy has been accepted as an effective and safe method to treat prostate cancer. The aim of this study was to describe acute toxicity following HDR brachytherapy to the prostate, and to examine the association between dosimetric parameters and urinary toxicity in low-risk prostate cancer patients. Material and methods Patients with low-risk prostate cancer were given HDR brachytherapy as monotherapy in two 12.5 Gy fractions. Planning objectives for the planning target volume (PTV) were V100% ≥ 90% and V150% ≤ 35%. Planning objectives for organs at risk were V75% ≤ 1 cc for the bladder, rectum and perineum, and V125% ≤ 1 cc for the urethra. Toxicity was assessed three months after treatment using the Common Terminology Criteria for Adverse Events. Results Seventy-three patients were included in the analysis. Thirty-three patients (45%) reported having any type of toxicity in the three months following HDR brachytherapy. Most toxicity cases (26%) were grade 1 urinary toxicity. Mean coverage index was 0.89 and mean V100 was 88.85. Doses administered to the urethra were associated with urinary toxicity. Patients who received more than 111.3% of the prescribed dose in 1 cc of the urethra were four times more likely to have urinary toxicity compared to patients receiving less than 111.3% (OR = 4.71, 95% CI: 1.43-15.6; p = 0.011). Conclusions High-dose-rate brachytherapy administered as monotherapy for prostate cancer proved to be a safe alternative treatment for patients with low-risk prostate cancer. Urinary toxicity was associated with the dose administered to 1 cc and 0.1 cc of the urethra and was remarkably inferior to the reported toxicity in similar studies. Purpose Prostate cancer is the leading cause of cancer incidence in males and the second cause of male cancer mortality in Colombia. In 2012, 9564 cases were diagnosed in the country [1]. Radiotherapy administered as either external beam radiotherapy (EBRT), high-dose-rate (HDR) interstitial brachytherapy or a combination of both modalities is a standard of treatment for prostate cancer. High-doserate brachytherapy was initially introduced as a boost after EBRT in the treatment of prostate cancer [2][3][4] and recommended by both European and American associations [5][6][7], particularly for patients with intermediate to high risk prostate cancer. In patients with low-risk prostate cancer, HDR brachytherapy as monotherapy is con-sidered as an alternative that could be administered in shorter periods of time, with similar efficacy, better dosimetric outcomes for organs at risk, and a lower probability of inter and intra-fractional displacements in contrast to EBRT [8]. The first studies that implemented interstitial HDR brachytherapy as monotherapy for prostate cancer used between eight and nine fractions in a five-day period [9]; afterwards, four fraction schemes were implemented in a two-day period [10]. These studies allowed HDR monotherapy to be accepted as an effective and convenient method to treat prostate cancer, providing a similar biochemical control of the disease, and low toxicity to organs at risk. This therapy continued to evolve into hypofractionation using two and even single doses [11][12][13]. These schemes proved to be convenient regarding costs and hospitalization days. Recent studies have demonstrated low toxicity and adequate local tumor control of two 12.5 Gy fractions applied in a single day as monotherapy for low-risk prostate cancer [7,9,11,12,14]. Clinical results in prostate cancer in Colombia have been published using permanent interstitial brachytherapy as monotherapy [15], and HDR brachytherapy either as an exclusive therapy applied in four fractions, or as a boost to EBRT applied in two fractions [16]. However, two-fraction HDR as monotherapy for prostate cancer is not yet a common practice in the country. The purpose of this study was to describe acute toxicity and examine possible associations between different dosimetric parameters and urinary toxicity in low-risk prostate cancer patients treated with exclusive HDR brachytherapy. Patients We conducted a retrospective chart review of all patients with low-risk prostate cancer (T1-T2a tumor, PSA ≤ 10 ng/ml and a Gleason score ≤ 6), who had been treated with HDR brachytherapy as monotherapy between August 2011 and January 2014. Patients were considered ineligible for the procedure if they had a history of transurethral resection of the prostate, an International Prostate Symptom Score

> 15, and were unable to assume the lithotomy position or had any contraindication to receive anesthesia. Implant procedure The procedure was performed under regional epidural anesthesia. All patients received ciprofloxacin as a prophylactic treatment. The implant was performed under trans-rectal ultrasound (TRUS) guidance using a 5 mm template and two fixation needles. Implant needles were placed 10 mm away from the urethra, and between 3-5 mm inside of the prostatic capsule in order to decrease the dose to the rectum. Seminal vesicles were not routinely implanted considering that all patients were low-risk cases. Two gold fiducial markers were implanted as a reference to verify needle position during treatment in orthogonal X-ray images. Implant needles were fixed to the template and the template was sutured to the perineum in order to reduce the probability of needle displacement between fractions. Brachytherapy was administered in two fractions of 12.5 Gy each, applied with a six-hour interval on the same day. Prior to administering brachytherapy at the second fraction, we verified needle position using orthogonal X-ray images, which were compared to the first fraction's set of images regarding needle position in relation to the gold fiducial markers. When needed, individual needles or the entire template were manually repositioned. If necessary, it was possible to reinsert the TRUS probe prior to the second application, however, this was not necessary for any patients and no treatments needed to be re-planned due to needle displacements. Volumedefinition,high-dose-rateplanning and dosimetric measures Clinical target volume (CTV) was defined based on the prostatic capsule without an extra margin. The volume of the urethra was defined using contrast media prepared by combining 15 cc of 2% lidocaine gel, 10 cc of saline solution, and air. This contrast was then applied via a urinary catheter. The rectum, bladder, and perineum were anatomically defined. Four auxiliary planning volumes to improve treatment optimization were defined: urethra + 4 mm, planning target volume (PTV) + 4 mm, body minus the PTV, and perineum ( Fig. 1). Inverse planning optimization based on anatomical volumes was performed using the Treatment Planning Software HDRplus 3.0 (Eckert and Ziegler BEBIG, Germany); manual optimization was employed as a complement to further improve dose coverage. Planning objectives for the PTV were V 100% ≥ 90% and V 150% ≤ 35%. Planning objectives for organs at risk were V 75% ≤ 1 cc for the bladder, rectum and perineum, and V 125% ≤ 1 cc for the urethra. Planning target volume coverage was reported using V 100 and D 90 . Planning target volume homogeneity was reported using V 150 and V 200 . Coverage index (CI), dose non-uniformity ratio (DNR), homogeneity index (HI), and conformality number (CN) were also reported. Dosimetry for organs at risk included the V 75 , V 115 , V 125 , as well as the D 1cc and D 0.1cc . Toxicity assessment According to our institution's treatment guidelines, patients attended an immediate control appointment during the first week after the procedure; follow-up visits were scheduled three months after treatment. Acute toxicity was evaluated during these visits. We included all events that occurred during the previous three months, even if these had resolved before the control visit. Common Terminology Criteria for Adverse Events version 4.03 [17] was used to evaluate and describe the proportion of patients presenting symptoms related to urinary, sexual or rectal acute toxicity, and to score its severity. Statistical analysis Acute urinary, rectal, and sexual toxicity were described using simple frequencies and proportions. Coverage, homogeneity, dosimetric indexes, and dosimetry for organs at risk were described using central tendency and dispersion measures. Associations between dosimetry and urinary toxicity were explored by several univariable logistic regression models, in which the dependent variable was the presence or absence of urinary toxicity, and the independent variables were several dosimetric indexes. Continuous numerical variables related to coverage, homogeneity, dosimetric indexes, and dosimetry for organs at risk were categorized into binary variables prior to their inclusion in the univariable model. Categorization into binary variables was performed based on the analysis of the receiver operating characteristic (ROC) curves produced for each dosimetric parameter, in order to identify the best cutoff point (the point showing the better compromise between sensitivity, specificity, percentage of correctly classified cases, and area under the curve [AUC]). Variables that did not have a clear cut-off point on the ROC curve were analyzed as continuous numeric variables. Odds ratios and 95% confidence intervals were obtained from each univariable logistic regression model. Wald tests were used to calculate p-values to test for the general association between each variable and urinary toxicity. All statistical analyses were carried out using STATA/SE version 12.1 (College Station, TX: StataCorp LP, USA). Results Between August 2011 and January 2014, a total of 92 patients with low-risk prostate cancer were treated with HDR brachytherapy as monotherapy for prostate cancer at our institution. Seventeen patients were excluded from the study because they did not attend the follow-up visit; another two patients were excluded because of lack of information on some dosimetric parameters; thus, the final analysis consisted of 73 patients. Mean age was 65.5 years, mean number of needles was 15.4, and the mean prostate volume was 44.9 cc. Thirty-three patients (45.2%) reported having any type of toxicity in the three months after receiving HDR brachytherapy. Most of the toxicity was grade 1 urinary toxicity (Table 1). Mean coverage index, mean D 90 , and mean V 100 showed a satisfactory coverage of the treatment volume. Mean HI, mean V 150 , and mean V 200 showed that heterogeneity was well controlled. Dosimetry for organs at risk showed that the planning objectives were achieved for most patients ( Table 2). The ROC analysis allowed us to categorize most of the numerical dosimetrical indexes ( Table 3). The indexes that showed the highest AUC were urethra V 115% with a cutoff point of 5.9%, urethra D 0.1cc with a cutoff point of 117.4%, and urethra D 1cc with a cutoff point of 111.3%. None of the bladder indexes showed a high AUC or a clear cut-off point. This analysis allowed us to decide on the best cut-off value for each variable before entering it in the regression model for analysis. Based on the logistic regression analysis, we found that the doses administered to the urethra were associated with urinary toxicity. Patients who received more than 111.3% of the prescribed dose in 1 cc of the urethra were four times more likely to have urinary toxicity compared to patients receiving less than 111.3% (OR = 4.71, 95% CI: 1.43-15.6; p = 0.011). Similarly, patients who received more than 117.4% of the prescribed dose in 0.1 cc of the urethra had a higher risk of urinary symptoms compared to those who received less than 117.4% (OR = 2.76, 95% CI: 1.00-7.63; p = 0.05) ( Table 4). Discussion High-dose-rate brachytherapy administered as monotherapy in two fractions of 12.5 Gy showed to be a safe treatment for patients with low-risk prostate cancer. This treatment alternative comprises advantages related with the reduction of hospitalization costs, caregiver, and administrative burden, as well as patient comfort. In comparison with low-dose-rate brachytherapy, HDR is a more economical alternative [18,19], with a greater potential of obtaining better dosimetric results, and with an even greater possibility to obtain dosimetric advantages regarding coverage, conformity, homogeneity, and dosage to healthy organs since the advent of inverse planning algorithms and optimization based in anatomical structures instead of geometric structures [20,21]. The observed grade 2 acute urinary toxicity in our study is remarkably lower than that reported in similar studies. Even though these studies used different scales to assess toxicity, we consider it to be a valid comparison when evaluating toxicity in different hypofractionation schemes. A similar finding is related to rectal toxicity, which was lower in our study in comparison with other studies (Table 5). However, sexual toxicity was higher in our study, although a caveat to this regard is that sexual toxicity was not routinely evaluated in the other series. The remarkably lower toxicities observed in our study could be related to a relatively low dose administered to the PTV, considering that the mean V 100% was 88.8%, and that even the 90 th percentile for V 100% was of 91.9%. Our coverage could be better considering that most of the studies using pre-planning and inverse planning optimi- zation based on anatomical volumes obtain mean V 100% values superior to 95%. It is possible that in the scenario of current practice, in which traditional dose constraints are more easily accomplished due to better technology, modified constraints could be considered such as urethra V 115 ≤ 6%, urethra D 1cc ≤ 110%, and urethra D 0.1cc ≤ 118% in order to decrease urinary toxicity. We obtained a decreased toxicity but with a coverage that could be better. However, D 90 and V 100 accomplished RTOG recommendations. Consequently, we consider that long-term outcomes related to biochemical control or survival free of metastases will not be compromised. In our practice, we will consider to increase PTV coverage, and further decrease the V 115 ≤ 6%, D 1cc ≤ 110%, and D 0.1cc ≤ 118% for the urethra. The study has a main limitation related to the size of the sample of patients, which is small; hence, the statistical power can only detect relatively big differences. The proportion of patients who did not attend the follow-up visit was considerable; this may have resulted in an underestimation of toxicity if the reason why patients did not attend was related to their disease. However, we did not evaluate why patients did not attend follow-up and are not able to draw any conclusions about loss to follow-up. The reduced sample size did not allow us to perform a multivariable analysis capable of controlling for confounding between the dosimetric indexes reported for urethra and some other variables such as prostate volume or heterogeneity. Future studies in our institution will consider the analysis of late toxicity and outcomes related with biochemical failure in a greater number of patients. The ROC analysis is regarded as a strength of the statistical analysis, since we were able to choose the most appropriate cut-off points before categorizing each numerical variable. Conclusions High-dose-rate brachytherapy as monotherapy for prostate cancer administered in two 12.5 Gy fractions proved to be a safe treatment alternative in the treatment of selected patients with low risk cancer, in the era of inverse planning optimization based on anatomical volumes. Future studies should consider increasing the PTV coverage when using inverse planning optimization based on anatomical volumes, and should consider limiting the doses administered to the V 115 , D 1cc and D 0.1cc in the urethra. A randomised controlled study on the effects of hernial sac stump fenestration on ultrasound seroma prevention in laparoscopic Type III inguinal hernia repair Background: The incidence of ultrasound seromas has significantly increased after large hernial sac surgery. Several methods are available for preventing ultrasound seromas, but the clinical results are poor. It has also been demonstrated that hernial sac stump fenestration during laparoscopic incisional hernia repair surgery can significantly decrease the incidence of ultrasound seromas. Materials and Methods: Ninety patients aged 18–75 years who were treated in our hospital for primary Type III indirect inguinal hernia from March 2017 to March 2018 were randomised to a preventive fenestration group and a control group. All patients underwent transabdominal preperitoneal repair. The number of ultrasound seromas in the inguinal regions and ultrasound seroma volume on day 6 and months 1 and 3 after surgery in the two groups were compared. The secondary outcomes included length of surgery, urinary retention, acute pain, chronic pain, length of hospitalisation, recurrence rate and other complications. Results: There were no significant differences in demographic characteristics. Ultrasound seroma incidence and ultrasound seroma volume on day 6 and months 1 and 3 after surgery were significantly lower in the preventive fenestration group than that in the control group. There were no significant differences in the length of hospitalisation or incidence of acute pain or urinary retention between the two groups. Conclusions: Hernial sac stump fenestration after hernial sac transection in inguinal hernia repair surgery is a simple method that can effectively reduce post-operative ultrasound seromas. INTRODUCTION Compared with open surgery, laparoscopic inguinal hernia repair surgery has advantages such

as mild post-operative pain, faster recovery and low recurrence rate. [1][2][3][4] However, post-operative ultrasound seromas are more common in these types of surgeries. Fan [5] reported that the incidences of ultrasound seromas 1 and 6 days and 1 month after laparoscopic inguinal hernia repair surgery were 25.6%, 60.3% and 13.2%, respectively. Li and Ruze [6,7] reported that the incidences of seromas 1 week and 1 month after laparoscopic inguinal hernia repair surgery were 12.9%-18.4% and 2.9%-14.5%, respectively, and the incidence of ultrasound seromas had significantly increased after a large hernial sac surgery. Small ultrasound seromas can be spontaneously absorbed, but larger ones can cause pain or even the illusion of hernial recurrence to both patients and physicians. This causes some patients to return to the hospital multiple times for diagnosis and treatment, thus decreasing patient satisfaction. [8] The prevention of ultrasound seromas after laparoscopic inguinal hernia repair surgery is a major problem for hernia surgeons. One study found that large hernial sac and scrotal hernial sac were the main factors for ultrasound seromas. Currently, there are many methods available to prevent ultrasound seromas (e.g., inserting a drainage tube in the preperitoneal space), but the clinical results are poor. [9] It has also been demonstrated that hernial sac stump fenestration during laparoscopic incisional hernia repair surgery can significantly decrease the incidence of ultrasound seromas. [10] The clinical effects of hernial sac stump fenestration during laparoscopic inguinal hernia repair surgery require further study. We conducted a randomised controlled study to compare the ultrasound seroma prevention effects between hernial sac stump fenestration and hernial sac stump non-treatment during laparoscopic inguinal hernia repair surgery. Study population This study was conducted at the Department of Gastrointestinal Surgery in the Second Affiliated Hospital of Kunming Medical University. Annually, the department conducts laparoscopic inguinal hernia repair surgery on 800 patients. Patients enrolled in this study were diagnosed with indirect inguinal hernia during consultation at the surgical outpatient clinic. The inclusion criteria were (1) male, (2) primary indirect inguinal hernia, (3) age 18-75 years and (4) Type III hernia according to the Gilbert classification. [11] The exclusion criteria were (1) direct inguinal hernia, irreducible hernia, recurrent hernia or incarcerated hernia; (2) history of lower abdomen surgery or peritonitis; (3) coagulation disorder; (4) severe psychiatric disorder; (5) chronic pain or long-term drug abuse and (6) significant comorbidities. The trial lasted from 1 March 2017, to 1 March 2018. The trial was approved by the Ethics Committee of the Second Affiliated Hospital of Kunming Medical University according to the Ethical Principles for Medical Research Involving Human Subjects in the Declaration of Helsinki published by the World Medical Association. All surgeries were conducted in accordance with good clinical practice and Chinese laws. All patients provided signed informed consent. All surgeries were completed by the same surgeon, who held an associate director title. Randomisation and blinding A computer-generated random sequence was used to randomise the two groups. Patients were randomised 1:1 to a preventive fenestration group and a control group (untreated hernial sac stump). During the post-operative follow-up period, the surgeon and the radiologist evaluated the clinical and ultrasound evidence for seromas, surgical outcomes and complications. The patients, the surgeon and the radiologist were blinded to treatment allocation. All data were collected prospectively. Study interventions The procedure for laparoscopic inguinal hernia surgery was carried out in strict accordance with the Operating Guidelines for Laparoscopic Inguinal Hernia Surgery stipulated by the Chinese Society of Surgery. In this study, after the Bogros space and Retzius space were freed, the hernial sac was transected at the inner hernial ring. Peritonealisation of the spermatic cord for 6-8 cm was performed, and the proximal end of the transected hernial sac was sutured and sealed, while the distal hernial sac was excluded. In the preventive fenestration group, after hernial sac transection [ Figure 1], the assistant pushed the bottom of the hernial sac out from the ring [ Figure 2] and used a dissecting forceps to lift the bottom of the hernial sac. Around 2 cm × 2 cm of hernial sac wall was resected from the bottom of the hernial sac using an electric scissors [ Figure 3], and attention was paid to haemostasis. The hernial sac was then reduced into the scrotal sac [Supplementary Video 1]. The hernial sac stump was not treated in the control group. EasyProsthes™ lightweight three-dimensional repair patches (Beijing, China), 12 cm × 16 cm, were used but not fixed. All surgeries were completed by the same surgical team. Seroma definition Ultrasound seroma is defined as a mass that does not increase in size with coughing and can be observed clinically. Further, the mass is confirmed as a fluid-containing structure using B-ultrasound at the inguinal region. Hernial size Hernial size = the distance from the inner ring to the distal end of the hernial sac × the transverse diameter of the hernia. Primary results The number and volume of ultrasound seromas at the inguinal region on day 6 and months 1 and 3 after surgery were compared between the two groups. Secondary results The secondary outcomes included length of surgery, length of hospitalisation, urinary retention, acute pain, chronic pain (pain persisting for 3 months), recurrence rate and other complications. Pain was measured using the visual analogue scale (VAS) (range: 0-10). A 10-cm line was drawn and marked equidistant, 0-10, with 0 meaning no pain and 10 meaning the most severe pain. Patients with a VAS >5 were considered to have acute pain. Sample size calculation and statistical analysis Sample size calculation was based on a retrospective analysis of hernial sac stump fenestration in inguinal hernia repair surgery from 1 June to 1 December, 2016. The incidence of seromas 6 days post-operatively after hernial sac stump fenestration was assumed to decrease by 50%. With a confidence interval of 95% and a power of 80%, the sample size per group was estimated to be at least 45. Prospective data such as demographics, hernia characteristics and hernia treatment were collected and analysed. Analysis was performed using SPSS version 18.0 (SPSS Inc., Chicago, IL, USA). All continuous variables were expressed as mean and standard deviation, and independent sample t-tests were used for analysis. The Chi-square test was used for all categorical variables. A difference of P < 0.05 was considered statistically significant. From 1 March 2017, to 1 March 2018, 125 consecutive patients were screened, among which 90 were ultimately enrolled and randomised. The main reasons for exclusion were refusal to participate in the study, direct inguinal hernia and significant comorbidities [ Figure 4]. All patients were male with Gilbert Type III hernia, had a hernial defect >3 cm and had undergone transabdominal preperitoneal repair. Seventy-seven patients had unilateral indirect inguinal hernia. There was no significant difference in hernial size before surgery between the two groups (23.6 ± 6.3 cm 2 vs. 22.8 ± 5.9 cm 2 , P > 0.05). There were no significant differences in demographic characteristics, such as age, sex, body mass index, hernial size and follow-up duration between the two groups [ Table 1]. Table 2 shows the surgical outcomes. The total incidences of ultrasound seromas on day 6 and months 1 and 3 after surgery were 32.2% and 16.7% and 6.7%, respectively. The incidences of ultrasound seromas on day 6 and months 1 and 3 after surgery in the preventive fenestration group were significantly lower than that in the control group (4.4% vs. 60.0%, P < 0.001 and 0 vs. 33.3%, P < 0.001 and 0 vs. 13.3%, P < 0.001, respectively). Ultrasound seroma volumes on day 6 and months 1 and 3 after surgery in the preventive fenestration group had significantly reduced compared with that of the control group (9.8 ± 2.5 ml vs. 32.5 ± 7.3 ml, P < 0.05 and 0 vs. 20.8 ± 4.5 ml, P < 0.05 and 0 vs. 15.7 ± 3.3 ml, P < 0.05, respectively). The length of surgery in the preventive fenestration group was longer than that in the control group (54.30 ± 8.5 min vs. 50.60 ± 7.30 min, P < 0.05). There was no significant difference in the number of patients with acute pain after laparoscopic hernia repair surgery or length of hospitalisation between the two groups. Chronic pain, early recurrence or other complications did not occur in the two groups. DISCUSSION Laparoscopic inguinal hernia repair surgery has many advantages, but ultrasound seromas have troubled clinicians and patients. In this study, we employed a simple technique to solve this problem. We conducted a randomised controlled trial to evaluate the effects of hernial sac stump fenestration on ultrasound seroma prevention after laparoscopic inguinal hernia repair surgery. In this study, the incidence of ultrasound seromas at day 6 after surgery was significantly lower in the preventive fenestration group compared with that of the control group (2 vs. 27), and ultrasound seromas spontaneously disappeared at the 1-month follow-up after surgery in those two patients. In the control group, there were 15 and 6 patients with ultrasound seromas at months 1 and 3 after surgery, respectively. The ultrasound seromas that were present at month 3 disappeared after multiple punctures and drainage. Ultrasound seroma volume in the preventive fenestration group at day 6 after surgery was significantly lower than that in the control group, and ultrasound seromas disappeared 1 month after surgery. In the control group, ultrasound seroma volumes at months 1 and 3 after surgery were 20.8 ± 8.1 ml and 15.7 ± 3.3 ml, respectively. The length of surgery was longer in the preventive fenestration group than that in the control group by 3.70 min, mainly because hernial sac fenestration increased the length of surgery. There were no significant differences in the incidence of post-operative acute or chronic pain, early recurrence or other complications between the two groups. These results showed that hernial sac stump fenestration can significantly reduce the incidence of post-operative seromas, does not increase complications and can increase patient satisfaction. During laparoscopic inguinal hernia surgery, tissue spaces caused by the stripping of the hernial sac are the main reason for ultrasound seromas. The surgical dissection wound continuously exudes liquid that flows into the distal hernial sac stump, which is lowest in the inguinal region. The hernial sac stump forms a closed cavity with inner ring adhesions, resulting in the ultrasound seromas that are commonly seen in clinical practice. [8] Ultrasound seromas mostly spontaneously disappear 2-3 months after inguinal hernia surgery, [12,13] but during that period, they cause anxiety, dissatisfaction and discomfort for patients and affect their normal life and work. In addition, ultrasound seromas in some patients do not completely disappear For small hernial sacs, there is a tendency for complete excision of the hernial sac. [14] For large or scrotal hernial sacs, complete stripping of the hernial sac is not only time-and-effort intensive but also increases the risk of spermatic cord injury and bleeding. Therefore, it is recommended that hernial sac transection be carried out. A certain proportion of patients will develop ultrasound seromas if the distal hernial sac is excluded but not treated after transection. [14,15] To prevent ultrasound seromas, Daes et al. [16] transected large hernial sacs before pulling the distal hernial sac from the scrotum and used tacks to fix these sacs on the ipsilateral anterior abdominal wall. Although this method can somewhat reduce the amount of fluid that enters the scrotum from the preperitoneal space and reduce ultrasound seromas, it increases surgical cost and duration, and a large proportion of patients still develop ultrasound seromas. Some studies [5,17,18] found that preperitoneal drainage could significantly reduce post-operative seromas, but this is difficult to apply in clinical practice for the following reasons: (1) fluid exudation from the wound on the preperitoneal space is a continuous process that usually persists for 1 week, (2) inserting a drainage tube causes discomfort for patients and increases the potential risk of infection and lastly, (3) the process prevents patients from being discharged earlier. The hernial sac stump fenestration technique used in this study is simple, can significantly reduce ultrasound seromas and has good results. The principles of this technique are that the fenestration of the bottom of the hernial sac is similar to the excision of tunica vaginalis, and there are abundant blood and lymphatic vessels in the spermatic cord and Dartos fascia that can rapidly reabsorb exudates. Hence,

fluid exudated from the surgical wound can be quickly absorbed by the spermatic cord and Dartos fascia after fenestration, thereby reducing ultrasound seromas. This study has two main limitations. First, it was a single-centre study with a small sample size; multicentric studies with large sample sizes are needed for validation. Second, we considered only Type III hernias and did not perform a detailed analysis of hernial size. CONCLUSIONS Preventive fenestration in laparoscopic inguinal hernia repair surgery is a simple, safe and effective method that can significantly reduce ultrasound seromas without increasing recurrence or the risk of acute or chronic pain. Traumatic Brain Injury in the Elderly: Clinical Features, Prognostic Factors, and Outcomes of 133 Consecutive Surgical Patients Objective With the aging of the global population, an increase in the proportion of elderly patients presenting with traumatic brain injury (TBI) is expected. This population presents several distinctive characteristics that impact management and outcome of TBI, such as comorbidities, frailty, and preinjury use of medications - specially antiplatelets and anticoagulants. The purpose of this study was to assess the general characteristics and prognostic factors of elderly patients with TBI that were surgically managed at a single institution. Methods The authors performed a retrospective review of all elderly patients (age ≥ 65 years) with a history of TBI that underwent cranial neurosurgical procedures at their institution, between 2015 and 2019. Clinical characteristics, laboratory tests, and radiological scans, as well as surgeries, performed, outcome, and prognostic factors were analyzed, comprising 133 consecutive cases overall. Results The mean age of patients was 76.6 ± 7.3 years, ranging from 65 years to 97 years. There was a predominance of males (71.4%) and the most frequent mechanism of injury was fall (80.4%). Mild TBI comprised 57.1% of the cases, followed by severe TBI in 25.6%. Frequent signs and symptoms were impaired consciousness (69.9%), focal motor deficits (32.3%), and gait disturbances (12.8%). The majority had reported comorbidities upon admission (79.7%), with cardiac disease (79.2%) and diabetes (24.5%) as the most frequent. Preinjury anticoagulation was reported in 18.8% and use of antiplatelet drugs in 17.3%. The most common finding in the head CT was chronic subdural hematoma (48.1%), followed by acute subdural hematoma (37.6%). Coagulation was found to be altered in 12.8% of the patients. The most common neurosurgical procedure performed was trephination for hematoma evacuation (56.3%), followed by craniotomy (21.2%). Blood product transfusion was needed in 61.7% of the patients. Overall mortality was 42.1%, with the majority in the first month after admission (83.9%). Unfavorable outcome (Glasgow Outcome Scale <5) at discharge was identified in 73% of the patients. Identified prognostic factors were TBI severity, absent pupillary reactivity, acute intracranial bleeding on head CT, basal cisterns obliteration, altered coagulation status, and need for blood transfusion. Conclusions TBI severity, pupillary reactivity, coagulation status, need for blood products transfusion and acute bleeding, as well as basal cisterns obliteration found in head CT, are factors that influenced the outcome in this series of elderly patients with TBI that need surgical management. It is paramount to observe the particularities of this population in this context, to optimize outcomes, avoid complications and ultimately generate awareness focused on prevention. Introduction It is currently well-known that the global population is aging, with estimates from the United Nations that one in six people in the world will be over 65 years by 2050 [1]. In the United States, it is expected that by 2034, the population older than 65 years will overpass the population under 18 years [2]. Likewise, in Brazil, it is estimated that 25% of the population will be over 65 years by 2060 -currently, this proportion is 9.2% [3]. 1 1 1 1 Traumatic brain injury (TBI) has three peak incidences, occurring mainly among children, young adults, and the elderly [4]. TBI alone results in more than 80,000 emergency department visits per year in patients aged 65 years or older in the United States. Out of this number, approximately 75% lead to patient hospitalization. And the leading causes of TBI in this population were falls (51%) followed by motor vehicle collisions (9%) and assaults (1%) [5]. Unlike their younger counterparts, the most common TBI manifestations in the elderly are contusions and subdural and intracerebral hemorrhage. The physiological changes that occur in this population explain the differences in the patterns of brain injury -reduced brain volume associated with increased subdural space, fragile walls of cerebral vessels -especially bridging veins -and adherence of the dura-mater to the inner surface of the cranium [6]. In addition, preexisting chronic medical conditions, such as diabetes, heart and kidney diseases, negatively impact survival in this population [7]. Also, preinjury use of anticoagulant and antiplatelet drugs poses a challenge for the management at the emergency department, dramatically increasing morbidity and mortality in the elderly [8][9][10][11][12]. The goal of this study was to analyze the clinical characteristics and prognostic factors of a consecutive series of elderly patients that underwent neurosurgical procedures at our institution following TBI. Population studied The authors performed a retrospective review of records and radiological scans of all elderly patients (age ≥ 65 years upon admission) with a diagnosis of TBI and who underwent any TBI-related neurosurgical procedure at the University Hospital of Ribeirão Preto Medical School, São Paulo, Brazil, a level I regional trauma center, between 2015 and 2019. Patients with radiological findings of chronic subdural hematoma, but no confirmed history of trauma were excluded from this study. All acute trauma patients were evaluated according to the Advanced Trauma Life Support guidelines of the American College of Surgeons, and further radiological and neurosurgical assessments were employed whenever deemed necessary by the admitting team. The remaining patients, for which there were no acute injuries (55% of total), were admitted to the neurosurgical ward for observation. Data were obtained from electronic medical charts and included demographic information, mechanism of injury, known comorbidities, status of anticoagulation and platelet antiaggregation, prescription drugs, clinical signs, and symptoms upon admission -including pupillary reaction -associated lesions, head computed tomography findings, laboratory results, surgeries performed, length of stay, complications, need for blood transfusion, outcomes (as measured by the Glasgow Outcome Scale -GOS -at discharge, 30 days, 60 days, six months, and one year after admission, if available) and mortality. Outcome was classified in favorable (GOS = 5) and unfavorable (GOS <5). The serum lactate and base deficit levels on admission were categorized as described by Callaway et al., for their known role as prognostic biomarkers; the lactate level was classified as normal (0 to 2.4 mmol/L), moderately elevated (2.5 to 4.0 mmol/L), or severely elevated (>4.0 mmol/L) and base deficit was classified as normal (> 0 mEq/L), moderate (0 to -6 mEq/L) or severe (< -6 mEq/L) [13]. Coagulation status was assessed by fibrinogen level and prothrombin time -expressed by international normalized ratio (INR) and classified as non-therapeutic (<2.0) or altered (≥ 2.0). Institutional approval for this study was obtained from the Ethics Committee of the University Hospital of Ribeirão Preto Medical School -University of São Paulo. Statistical analysis GraphPad Prism 8 (version 8.4.2, GraphPad Software, San Diego, CA) was used for the statistical analyses. Descriptive data were presented as means (with standard deviation), median -when applicable, and proportions. Univariate analysis was performed with Fisher's exact test for categorical variables, except for admission Glasgow Coma Scale, lactate level and base deficit, which were analyzed with chi-square. Oddsratio (OR) with 95% confidence intervals (CI) were calculated. Statistical significance was assumed when ɑ = 0.05. Characteristic Value Age ( Discussion The main mechanisms of injury in this series were fall (80.4%), motor vehicle collision (11.3%) -pedestrian and vehicle occupant -and assault (1.5%), but no gun-related injuries. However, there was a predominance of fall over the remaining mechanisms, when comparing with other reported series (30-51% of trauma cases in this age group) that account for elderly patients with TBI in general -surgical and conservative cases [8,14,15]. This may be explained by previously described studies depicting the higher injury severity and increased need for surgical management of elderly patients that present with this type of mechanism [16]. Clinical presentation was variable, however, 69.9% of the patients had an impaired consciousness upon admission, followed by focal motor deficits and gait disturbances. This raises the importance to alert caretakers and family members for changes in behavior, focal weakness, and gait disturbances, signs that may be overlooked and mistakenly attributed to physiological consequences of aging, leading to underreporting and failure to seek medical attention promptly [17]. Also, an episode of fall may be unnoticed or the TBI ignored by caretakers or medical staff because of distracting lesions, such as orthopedic trauma [18]. A large study assessing the influence of preexisting medical conditions in elderly trauma patients identified 33% of patients with at least one comorbidity and it was associated with an increased risk of death [7]. In this study, approximately 80% of the patients had at least one reported chronic disease upon admission, thus suggesting a higher risk of surgically managed TBI in this population. Also, it is important to consider the prevalence of acute and chronic abusive alcohol drinking in this population. It is reported that in elderly patients 13% of men and 8% of women reported at-risk alcohol use, with respectively 14% and 3% of binge drinking [19]. In our series, chronic abusive alcohol use was reported in 6.6% of the patients. The role of polypharmacy (i.e., the use of multiple medications and/or the administration of more medications than are clinically indicated, representing unnecessary drug use) spanning from inappropriate use of prescribed medicines, as well as self-medicating without an indication, in this age group is associated with the high prevalence of comorbidities [20]. Preinjury use of anticoagulant or antiplatelet drugs is an example of a worrisome issue in this population, due to the higher risk of complications and mortality following low energy traumas, such as falls. Besides the prompt identification of the coagulation status of the elderly patient with TBI, is the need for urgent reversal of anticoagulation or antiplatelet agents, when a TBI is suspected, in order to reduce the risk of mortality or unfavorable outcome [21][22][23][24]. Although 18.8% of the patients in this series were reported to use anticoagulation medicine previous to admission, only the coagulation status assessed by prothrombin time (≥2.0) was found to significantly affect mortality, thus reinforcing that anticoagulation per se doesn't affect the outcome, but the anticoagulation status does [25]. The high mortality and predominant proportion of unfavorable outcomes in this series underscore the importance of the already reported specific protocols for the management of elderly patients with TBI. Protocols spanning from hospital admission, inpatient care -if possible, in a dedicated unit or early admission to an intensive care unit -and transfer to a rehabilitation facility or discharge with an integral follow-up [26][27][28][29][30]. Also, it is paramount to address the prevention of this type of trauma, by enrolling interventions focused both in the host/agent and the environment, in all the phases regarding the event -for example, in the pre-event phase addressing polypharmacy, improving strength and balance and eliminating environmental hazards [5]. Limitations The major limitation of this study is that it only comprises elderly patients with TBI that underwent any neurosurgical procedure following admission. This limitation is due to the nature of the database used -the registry of all neurosurgical procedures of our institution. Thus, it is expected that the results will be impacted towards worse clinical presentations, higher mortality, and unfavorable outcomes. Also, patients that had a favorable outcome upon hospital discharged, were commonly discharged from the neurosurgery outpatient clinic if no complications or limitations were detected, thus reducing the data available for outcome measures at 60 days, six months, and one-year post-admission. Another limitation may be that not all the patients had been tested for fibrinogen, serum lactate, and base deficit levels upon admission -in our institution only the severe and acute trauma patients have the blood gases and fibrinogen tested, thus reducing the statistical power of the analysis of these factors. received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to

have influenced the submitted work. The Homeodomain Transcription Factor Hoxa2 Interacts with and Promotes the Proteasomal Degradation of the E3 Ubiquitin Protein Ligase RCHY1 Hox proteins are conserved homeodomain transcription factors known to be crucial regulators of animal development. As transcription factors, the functions and modes of action (co-factors, target genes) of Hox proteins have been very well studied in a multitude of animal models. However, a handful of reports established that Hox proteins may display molecular activities distinct from gene transcription regulation. Here, we reveal that Hoxa2 interacts with 20S proteasome subunits and RCHY1 (also known as PIRH2), an E3 ubiquitin ligase that targets p53 for degradation. We further show that Hoxa2 promotes proteasome-dependent degradation of RCHY1 in an ubiquitin-independent manner. Correlatively, Hoxa2 alters the RCHY1-mediated ubiquitination of p53 and promotes p53 stabilization. Together, our data establish that Hoxa2 can regulate the proteasomal degradation of RCHY1 and stabilization of p53. At the molecular level, the transcriptional activities of Hox proteins have been well documented and a handful of Hox transcriptional cofactors could be identified, such as Pbx or Meis proteins that belong to the TALE proteins family [14,15]. Hox transcriptional targets were also functionally identified through transcriptomic screenings [16,17] and chromatin immunoprecipitation assays [18][19][20][21]. In parallel to their transcriptional activity, some Hox proteins have been involved in non-transcriptional processes. First, Hox proteins have been associated with translational functions. HOXA9 and HOXA13, for example, interact with the initiation factor of translation, eIF4E [22,23]. Interaction between HOXA9 and eIF4E stimulates mRNA transport by eIF4E and translation efficiency [22]. Second, Hox proteins have been involved in DNA double-stranded break (DSB) repair. HOXB7, for instance, interacts with two DNA repair proteins, Ku70/80, involved in nonhomologous end joining pathway (NHEJ) of DSB repair and confers resistance to DNA damage on irradiated cells [24]. Third, homeodomain proteins can cross biological membranes [25,26]. This feature, coupled with the possibility to be secreted into the extracellular milieu might confer to homeodomain proteins a direct cell-signalling activity. Finally, Hox proteins have been shown to play a crucial role in cell cycle regulation through the control of DNA replication. HOXD13, for example, binds DNA replication origins, primarily during G1 phase of the cell cycle, promotes the assembly of pre-replication complexes and induces DNA synthesis [27]. Such a role in cell cycle has also been suggested for HOXC10 [28] or Hoxb4, which has been associated with hematopoietic cell proliferation [29]. In this context, Hoxb4 takes part in an E3 ubiquitin ligase complex that specifically recognizes Geminin, an anti-replicative protein, and induces its degradation [30]. Therefore, it appears that Hox proteins may not simply be active as transcription factors. However, evidence about the molecular activities of Hox proteins in a non-transcriptional context remain sparse and need more extensive investigation. Hoxa2 has been involved in rostral hindbrain and neural crest patterning. Knock-out mice for Hoxa2 display defects in hindbrain segmental identity affecting the second and third rhombomeres [31][32][33]. At this level, Hoxa2 has been shown to be crucial for axon guidance and the building of sensorimotor circuitry connecting the brainstem to upper nervous centres [33,34]. Hoxa2 is also essential for the identity of neural crest cells migrating from the hindbrain to the second branchial arch, which participate to the formation of skeletal derivatives notably within the middle ear [31,32]. At the molecular level, some Hoxa2 transcriptional targets such as Lmo1, Meox1, Lhx6, Ptx1, Robo2 or Six2 have been reported [35][36][37][38][39]. However, only Six2, Robo2 and Hoxa2 itself, have been characterized so far as direct Hoxa2 target genes [35,[39][40][41]. A genome-wide survey of Hoxa2-bound sequences in the second branchial arch correlated to transcriptomic analyses allowed identifying large sets of candidate genes under the immediate control of Hoxa2 [18]. As interaction partners, only TALE proteins could be directly related to Hoxa2 molecular function [40] and a few candidate Hoxa2-interacting proteins have been identified mainly in the context of highthroughput experiments [42]. At the cellular level, it has been reported that Hoxa2 inhibits cell differentiation in the chondrogenic context and seems to be involved in cell migration [43][44][45]. Hoxa2 also inhibits differentiation of oligodendrocytes and, in addition, promotes their proliferation [46,47]. Conversely, Hoxa2 has been proposed to display an anti-proliferative activity during lung development [48]. However, the functional connection between molecular targets, cellular activities and developmental roles of Hoxa2 remains largely to be unveiled, as it is the case for other Hox proteins. In this study, we identified three unexpected interaction partners for Hoxa2: the RING finger and CHY zinc finger domain-containing protein 1, RCHY1 (also known as PIRH2), and 20S proteasome subunits PSMA3 and PSMB2. RCHY1 is an E3 ubiquitin ligase that has been mainly involved in cell cycle and apoptosis through its activity towards the key cell cycle regulators p53 and p27Kip1 [49,50]. As a consequence of the interaction between Hoxa2 and RCHY1, we further show that Hoxa2 specifically triggers the degradation of RCHY1 in an ubiquitin-independent and proteasome-dependent manner, and in turn stabilizes p53 protein level. Together, our results show that Hoxa2 is involved in negative regulation of the RCHY1 E3 ubiquitin ligase and identify a new molecular pathway connecting Hox proteins to the p53 protein homeostasis. Hoxa2 interacts with RCHY1, PSMA3 and PSMB2 To gain an insight into the mode of action of Hoxa2, we performed a stringent high-throughput GAL4-based yeast twohybrid screen optimized for the testing of the entire human ORFeome [51,52]. The human ORFeome v3.1 consists of an extensive set of cloned open reading frames derived from the human genome and corresponding to 10,214 human protein coding genes [52]. Using Hoxa2 both as a bait and prey, we screened the 12,212 open reading frames (ORFs) of the human ORFeome v3.1 and identified a short isoform of the RING finger and CHY zinc finger domain-containing protein 1 (RCHY1, better known as PIRH2), the proteasome subunit alpha type-3 (PSMA3) and the proteasome subunit beta type-2 (PSMB2) as Hoxa2-interacting proteins. These results were then confirmed by retransforming expression vectors for these potential partners into yeast and retesting them against Hoxa2 ( Figure 1A). To validate these results by co-precipitation experiments in mammalian cells, we constructed expression vectors for Nterminally triple-FLAG-tagged Hoxa2 (FLAG-Hoxa2) and Nterminal glutathione S-transferase (GST)-PSMA3, -PSMB2 or -RCHY1 (full length) fusion proteins. These vectors were then co-transfected in HEK293T cells and glutathione-agarose beads were used to co-purify GST-fused interactors and FLAG-Hoxa2. As a positive control, we chose to use Hoxa1 dimer formation which had already been reported by co-precipitation [53,54] ( Figure 1B). Cells transfected with FLAG-Hoxa2 alone were used as negative control. In the absence of expression vector for GST-tagged protein, no or weak background binding of FLAG-Hoxa2 was detected ( Figure 1B). However, we successfully retrieved FLAG-Hoxa2 from cells co-expressing GST-PSMA3 and GST-PSMB2, therefore validating these proteins as Hoxa2 interaction partners ( Figure 1B). Conversely, we failed to recover FLAG-Hoxa2 upon GST-RCHY1 coexpression (data not shown). While verifying that the fusion proteins were properly expressed in transfected cells, we were surprised to observe that, compared to cells transfected with the GST-RCHY1 vector alone, cells co-transfected with FLAG-Hoxa2 expression vector showed barely detectable weak GST-RCHY1 protein levels. This therefore indicated that expression of Hoxa2 had a negative effect on GST-RCHY1 protein accumulation. Since the GST-RCHY1 expression construct was based on a constitutively active CMV promoter, we hypothesized that the influence of Hoxa2 on the RCHY1 protein level was most likely due to an impact on RCHY1 protein stability. Hoxa2 induces RCHY1 degradation in a proteasomedependent way Since we suspected that RCHY1 protein stability was affected in presence of Hoxa2, we used a proteasome inhibitor, MG132, as the proteasome mediates one of the two main pathways of intracellular protein degradation. We exposed HEK293T cells to 1 µM of MG132, or to DMSO alone as a negative control, 24 hours after transfection, during 15 hours. In these conditions, high protein levels were observed for both FLAG-Hoxa2 and GST-RCHY1 in MG132 treated cells ( Figure 1C). We next performed affinity co-purification assays using glutathione-agarose beads and validated RCHY1 as an interaction partner for Hoxa2 ( Figure 1C). The striking observation regarding RCHY1 protein levels in these assays focused our attention on the degradation pathway of RCHY1. To confirm the suspected involvement of Hoxa2 in Hoxa2 and RCHY1 PLOS ONE | www.plosone.org Yeast transformed with expression vectors for GAL4-DB and GAL4-AD fusion proteins were mated on complete medium (YPD) and transferred on synthetic dropout medium plates lacking histidine, leucine and tryptophan (-L-W-H) to select diploids in which the GAL1-HIS3 reporter is activated as a consequence of hybrid proteins interaction. Negative control plates were composed of synthetic dropout medium containing cycloheximide and lacking histidine and leucine (-L+W-H+C) and positive controls for matings were transferred on synthetic dropout medium plates lacking leucine and tryptophan (-L-W). (B) Co-precipitation assays. HEK293T cells were co-transfected with expression vectors for FLAG-and GST-tagged Hoxa1 as a positive control, for FLAG-tagged Hoxa2 alone as a negative control and for FLAG-tagged and GST-tagged PSMA3 or GST-tagged PSMB2. Forty-eight hours after transfection, cell lysates were subjected to western blotting analyses to detect protein expression (beta-actin used as a protein load control). Protein interactions were then verified by coprecipitation on glutathione beads directed toward the GST tag. Eluted proteins were analysed by western blotting using M2 antibody to detect the presence of FLAG-tagged Hoxa2 (CoP). (C) Similar co-precipitation assays using MG132 or DMSO treated HEK293T cells reveal FLAG-tagged Hoxa2 and GST-tagged RCHY1 interaction upon proteasome inhibition. doi: 10.1371/journal.pone.0080387.g001 RCHY1 proteasomal degradation, we co-transfected vectors for wild-type Hoxa2 and FLAG-tagged RCHY1 proteins in HEK293T. Cells were then treated with 1µM of MG132, or DMSO, during 15 hours. As shown in Figure 2A, when wildtype Hoxa2 and FLAG-tagged RCHY1 expression vectors were co-transfected, a severe depletion in FLAG-RCHY1 protein could be observed in comparison with the accumulation of FLAG-RCHY1 when it is expressed alone (Figure 2A). In addition, this Hoxa2-associated reduction in FLAG-RCHY1 protein levels was rescued by MG132 treatment thereby confirming a proteasome-dependent degradation of RCHY1 ( Figure 2A). As the primary function of RCHY1 is to be an E3 ubiquitin ligase, it has also to be noticed that, contrary to what one might expect consequently to its interaction with Hoxa2, no significant decrease in Hoxa2 protein level could be observed upon co-expression with FLAG-RCHY1 ( Figure 2A). Proteasomal degradation occurs both in the nucleus and cytoplasm. As RCHY1 can be found in both subcellular compartments [55,56], the next question was to determine if we could observe a change in subcellular localization upon coexpression with Hoxa2. HEK293T cells were co-transfected with expression vectors for FLAG-Hoxa2 and GST-RCHY1 and treated with 1µM of MG132 for 15 hours. Cells were then processed to detect FLAG-and GST-tagged proteins and observed under confocal microscopy to characterize protein subcellular localization. As expected, when expressed alone GST-RCHY1 was present both in the nucleus and cytoplasm and it was more abundant when cells were treated with MG132 than in untreated cells ( Figure 2B). A classical nuclear staining for Hoxa2 and a reduced staining for RCHY1 were observed in untreated cells co-expressing FLAG-Hoxa2 and GST-RCHY1 ( Figure 2B). Finally, corresponding MG132-treated cells presented a nuclear co-localization for Hoxa2 and RCHY1, and no significant change in their respective intracellular distribution could be detected ( Figure 2B). Since Hoxa2 and RCHY1 colocalize mainly in the nucleus, it can be assumed that their interaction, and the subsequent RCHY1 degradation, takes place in this subcellular compartment. Hoxa2-induced RCHY1 degradation is ubiquitin independent As poly-ubiquitination is the main pathway for proteasomal degradation, and the only one currently known for RCHY1 degradation [49], we addressed the possibility that Hoxa2 triggers an ubiquitin-dependent RCHY1 decay. To test this, we co-transfected HEK293T cells with expression vectors for 6Histagged ubiquitin octamer and FLAG-RCHY1, with or without expression vector for wild-type Hoxa2. Again, cells were treated with 1µM of MG132 for 15 hours to inhibit the proteasome. Protein lysates were then purified for 6Hisubiquitinated proteins using NiNTA-beads and attached proteins were then loaded on SDS-PAGE to detect RCHY1 poly-ubiquitinated forms. Unexpectedly, in the presence of Hoxa2, a reduced RCHY1 poly-ubiquitination profile was observed, indicating that the Hoxa2-induced RCHY1 degradation is ubiquitin-independent. Moreover, as less RCHY1 ubiquitinated forms were detected in the presence of Hoxa2, this suggests that the poly-ubiquitination of RCHY1 is altered upon Hoxa2 interaction ( Figure 2C). These results were further confirmed with GST-RCHY1 (data not shown). The

Hoxa2 homeodomain is essential for the Hoxa2mediated RCHY1 degradation To further provide mechanistic insights into the Hoxa2mediated RCHY1 degradation, we performed experiments using mutant Hoxa2 proteins. A first Hoxa2 mutant harbours amino-acid substitutions in its homeodomain (Hoxa2 KQN-RAA ). In this mutant, the substituted glutamine and asparagine define two critical amino acid residues involved in the Hox-DNA interactions established by the third helix of the homeodomain which are shared by all Hox proteins [57]. The Hoxa2 KQN-RAA protein is therefore DNA-binding defective [40]. A second mutant displays substitutions in the short hexapeptide motif involved in the interaction of Hoxa2 with the Pbx proteins (Hoxa2 WM-AA ). As previously shown, these Hoxa2 variants are transcription defective [40]. Co-precipitation assays from MG132 treated HEK293T cells transfected for Flag-Hoxa2, Flag-Hoxa2 KQN-RAA or Flag-Hoxa2 WM-AA and GST-RCHY1 or a GST control revealed that both mutants were still able to bind to RCHY1 ( Figure 3A). However, while Hoxa2 and Hoxa2 WM-AA GST fusion were prominent in inducing degradation of Flag-RCHY1, the homeodomain mutant was not. Co-expression of Flag-RCHY1 and either GST-Hoxa2 or GST-Hoxa2 WM-AA resulted in the disappearance of the Flag-RCHY1 in untreated cells where the proteasome was active, whereas upon expression of GST-Hoxa2 KQN-RAA , the level of Flag-RCHY1 was unaffected ( Figure 3B). Since Hox proteins share important sequence similarities, in particular at the level of the homeodomain, we tested whether another Hox protein, namely Hoxa1, also displayed the ability to interact with and provoke the degradation of RCHY1. Coprecipitation from HEK293T transfected cells expressing Flag-Hoxa1, and GST-RCHY1 or a GST control supported that Hoxa1 shares the ability to bind to RCHY1 ( Figure 4A). Remarkably however, unlike Hoxa2, expression of a Hoxa1 fusion protein did not affect the level of RCHY1 in either MG132 or untreated cells ( Figure 4B). This supports that although able to contact RCHY1, Hoxa1 is not proficient in targeting RCHY1 to proteasomal degradation. Hoxa2 inhibits RCHY1-dependent ubiquitination of p53 and stabilizes p53 protein level p53 turnover has recently been revealed to occur through ubiquitination by RCHY1 and subsequent proteasomal degradation [49]. Since we observed that Hoxa2 can induce RCHY1 degradation, to get an insight into the possible biological consequences of the Hoxa2-RCHY1 interaction, we next examined the effect of Hoxa2 on p53 ubiquitination. We co-transfected cells with expression vectors for FLAG-RCHY1, p53 R72 variant (dbSNP: rs1042522) and 6His-tagged ubiquitin octamer, with or without a vector for wild-type Hoxa2. Cells were then treated with MG132 and protein lysates were purified using NiNTA-beads. Purified proteins were loaded on a SDS-PAGE gel to detect p53 poly-ubiquitinated forms. In the presence of Hoxa2, we observed less p53-ubiquitinated forms indicating that Hoxa2 negatively affects RCHY1-dependent p53 Hoxa2 and RCHY1 PLOS ONE | www.plosone.org HEK293T cells were cotransfected with FLAG-tagged Hoxa2 and GST-tagged RCHY1, treated with MG132 proteasome inhibitor, or DMSO, and subjected to immunocytochemistry with anti-FLAG M2 antibody (green) and anti-GST antibody (red). Nuclei were stained with DAPI (blue). (C) Ubiquitination assays for FLAG-tagged RCHY1. HEK293T cells were co-transfected with indicated expression vectors and treated with MG132. Cells were then lysed and 6His-ubiquitin conjugated proteins were purified using Ni-NTA beads. Purified proteins and cell lysates were analysed by western blotting using M2 antibody to detect ubiquitinated forms of FLAG-tagged RCHY1. Lysate samples were loaded on a SDS-PAGE to verify protein levels prior to purification (Input; beta-actin was used as a protein load control). Lane numbering under the gels identifies cell samples. I: input sample; P: Ni-NTA purified sample. (FLAG-Hoxa2KQN-RAA), FLAG-tagged Hoxa2 WM-AA (FLAG-Hoxa2WM-AA), GSTtagged RCHY1 and GST proteins, and treated with the proteasome inhibitor MG132. Forty-eight hours after transfection, cell lysates were subjected to western blotting (input) and protein interactions were verified by co-precipitation on glutathione beads directed toward the GST tag. Eluted proteins were analysed by western blotting using the M2 anti-FLAG antibody (CoP). (B) Amino acid substitutions in the Hoxa2 homeodomain abolish the Hoxa2-mediated degradation of RCHY1. HEK293T cells were transfected with expression vectors for FLAG-tagged RCHY1 (FLAG-RCHY1) and GST-tagged Hoxa2 (GST-Hoxa2wt, GST-Hoxa2KQN-RAA, GST-Hoxa2WM-AA) proteins. Cells were then treated for proteasome inhibition (MG132) and compared to untreated controls (DMSO) for the decay of FLAG-RCHY1 revealed by western blot detection. Detection of beta-actin was used as a protein load control. ubiquitination in MG132 treated HEK293T cells ( Figure 5A). As cells were treated with MG132, these results suggest that Hoxa2 inhibits RCHY1 ubiquitin ligase activity as such. To further investigate the effect of Hoxa2 on p53 stabilization, we next transfected wild-type Hoxa2 expression vector in PA1 cells. In the presence of Hoxa2, we observed a stabilization of endogenous p53 protein level ( Figure 5B). Together, these results indicate that Hoxa2 negatively affects RCHY1 stability and RCHY1-dependent ubiquitination of p53 thereby inducing a stabilization of p53 and an increase in p53 protein level. Discussion While Hox proteins are well established transcription factors acting as cornerstones in the regulation of developmental processes or in some instances of oncogenesis, the molecular interactions underlying their functions have been rather poorly investigated and have been basically focused on their transcriptional activity. Nevertheless, a growing body of evidence supports that Hox proteins could be involved in other processes like translational regulation, DSB repair or DNA replication [22,24,27,30]. Here, we have characterised a new role of Hoxa2 in the proteasomal protein degradation pathway. We identified the RING finger and CHY zinc finger domaincontaining protein 1, RCHY1, and two 20S core proteasome subunits, PSMA3 and PSMB2, as novel interaction partners for Hoxa2. We further showed that Hoxa2 induces proteasomedependent RCHY1 degradation, independently of the ubiquitin system, and promotes p53 protein stabilization. We thus provide evidence for a new molecular mechanism that possibly links Hox protein activity to p53-related pathways. Our results show that Hoxa2 interacts with RCHY1, an E3 ubiquitin ligase involved in specific degradation of key cell cycle regulators such as p53 or p27Kip1 [49,50]. This interaction promotes proteasomal degradation of RCHY1 and we provide data supporting that this degradation takes place independently of the ubiquitin system. Such a proteasomal degradation pathway independent of the ubiquitin system has already been reported for various proteins and, for some of them, the activity of the isolated 20S proteolytic core particle has been directly involved in the mechanism [58][59][60]. Similarly, our results show interaction between Hoxa2 and some subunits of the 20S core particle such as PSMA3, for example, which has been directly involved in the ubiquitin-independent degradation of p21Cip1 or Rb [58,61], indicating that a similar mechanism could be involved in Hoxa2-mediated degradation of RCHY1. Since we observed a nuclear co-localization of Hoxa2 and RCHY1 upon proteasome inhibitor treatment, we propose that Hoxa2 interacts with 20S core proteasome to directly induce nuclear RCHY1 proteasomal degradation in an ubiquitin-independent manner. Distinct proteins had already been proposed to be involved in the turnover of RCHY1 [56,62,63]. However, despite its crucial role in cancer and other pathologies [64][65][66], only two physiological factors have been reported to negatively regulate RCHY1 at the protein level: RCHY1 itself [49] and calmodulindependent protein kinase II (CaMKII) [63]. Indeed, RCHY1 may be self-ubiquitinated or subject to posttranslational Hoxa2 and RCHY1 PLOS ONE | www.plosone.org phosphorylation by CaMKII both leading to a decrease in RCHY1 protein levels [49,63]. Here, we thus identify Hoxa2 as a new factor involved in negative RCHY1 protein regulation. In addition to the enhanced turnover of RCHY1 by Hoxa2, we also observed a decrease in RCHY1 and p53 ubiquitination profiles in the presence of Hoxa2. This suggests that Hoxa2 interaction with RCHY1 also inhibits overall RCHY1 ubiquitin ligase activity. At the molecular level, the ability to promote RCHY1 degradation (while not its binding) relies on the integrity of the homeodomain. However, although necessary to induce RCHY1 decay, the homeodomain determinants should not be sufficient. Indeed, although sharing a highly conserved homeodomain, the Hoxa1 protein could not stimulate RCHY1 decrease. HEK293T cells were co-transfected with indicated plasmids and treated with MG132. Cells were lysed and Ni-NTA beads were used to pull down 6His-ubiquitin-conjugated proteins. Proteins were separated by SDS-PAGE and western blotting was performed using an anti-p53 antibody to detect ubiquitinated forms of p53. Lysate samples were loaded on a SDS-PAGE to verify protein levels prior to purification (Input; beta-actin was used as a protein load control). Lane numbering under the gels identifies cell samples. I: input sample; P: Ni-NTA purified sample. (B) p53 protein stabilization by Hoxa2. PA1 cells were transfected with indicated plasmids and treated with MG132, or DMSO as control. Cells were lysed, proteins were separated by SDS-PAGE and western blot detection was performed with indicated antibodies. Our results support a model in which Hoxa2 acts as a stabilizer for p53 protein level, via its negative effect on RCHY1. This could not be expected from the known roles of Hoxa2. Indeed, p53 is known to be mainly involved in cell cycle arrest, senescence and apoptosis [67]. Thus, a p53 stabilization would, at a first glance, induce cell cycle arrest, senescence and/or increase in apoptosis, which is not in line with Hoxa2 functional studies that led to postulate antidifferentiation and pro-proliferative roles for Hoxa2 [43,46,68]. Nevertheless, besides its role in cell cycle and apoptosis, p53 has also been involved in DNA repair. The p53 response to DNA damage varies according to its subcellular localization, the cell cycle status and the extent of DNA damage, from apoptosis induction to DNA repair [69]. Efficient DNA repair is a crucial requirement during embryonic development [70]. It allows cells sustaining a high division rate, implying shorter cell cycles, but preserving quality divisions essential for developmental processes. In that context, analyses with the KEGG PathwayFinder module from the R2 microarray analysis and visualization platform (http://r2.amc.nl) allowed highlighting a wide range of transcripts positively correlated to HOXA2 expression which correspond to genes involved in DNA repair (data not shown). In support of this hypothesis, RCHY1 has been shown to induce the degradation of PolH, a member of the Y family translesion DNA polymerase involved in doublestrand break repair via homologous recombination [71,72]. In addition, the study of the knock-out mouse for Rchy1 has recently been published and confirmed its role in DNA repair [73]. Molecular activities related to such basic cell processes as DNA repair or protein turnover have already been described for Hox proteins of distinct paralog groups [24,30]. Hoxb4 has been involved in DNA replication through a direct involvement in an E3 ubiquitin ligase complex that targets Geminin for degradation [30]. HOXB7 has been directly related to DNA repair processes [24]. Considering the general involvement of Hox proteins in the control of developmental processes and their functional redundancy, it is tempting to propose that the implication of Hox proteins in DNA repair processes and protein degradation pathways could be more general than just the matter of Hoxa2. Finally, while mutant Hoxa2 proteins assayed for their ability to induce RCHY1 degradation still appear to interact with RCHY1, amino acid substitutions in the Hoxa2 homeodomain critically impairs its impact on RCHY1 turn over. A similar observation was reported for the involvement of Hoxb4 in the E3 ubiquitin ligase complex active towards Geminin. A single amino acid substitution in the Hoxb4 homeodomain was sufficient to impair its ability to enhance the poly-ubiquitination activity although the mutant protein was still able to form the complex [30]. Conclusions In conclusion, we have demonstrated that the 20S proteasome subunits PSMA3 and PSMB2 as well as the E3ubiquitin-ligase RCHY1 are direct interactors of Hoxa2. We further showed that Hoxa2 promotes the proteasomedependent and ubiquitin-independent degradation of RCHY1 which in turn is correlated to p53 stabilization. The RCHY1 degradation stimulated by Hoxa2 requires the integrity of its homeodomain. Plasmid constructs Expression vector for wild-type Hoxa2 was described previously [38]. Sequences coding for the wild-type Hoxa2, and mutant Hoxa2 KQN-RAA and Hoxa2 WM-AA proteins [40] were PCRamplified and inserted into pDON223 vector using the Gateway® Technology from Invitrogen. The resulting entry plasmids were confirmed by DNA sequencing. Entry vector was then used to generate yeast expression vectors for ADand DB-tagged Hoxa2 with pDEST-AD and pDEST-DB destination vectors (Gateway®, Invitrogen); mammalian expression vectors for FLAG-tagged Hoxa2, Hoxa2 KQN-RAA and Hoxa2 WM-AA , with v1899 destination vector (for a N-terminal triple FLAG-tag fusion [74]); and mammalian expression vectors for GST-tagged Hoxa2, Hoxa2 KQN-RAA and Hoxa2 WM-AA proteins (pDEST-GST N-terminal [75]). Entry vectors for PSMA3, PSMB2 and full length RCHY1 are from the hORFeome v3.1 [76] and

were used to generate destination mammalian expression vectors for N-terminal GST fusion proteins (pDEST-GST N-terminal [75]). The coding sequence for RCHY1 was also transferred into v1899 destination vector to produce a FLAG-tagged RCHY1. Expression vectors for Nterminal triple FLAG and N-terminal GST fusion Hoxa1 have been described elsewhere [54]. Expression vectors for 6Histagged ubiquitin and p53R72 were kindly offered by Sonia Lain (University of Dundee, UK) and Patrick Dumont (Université catholique de Louvain, Belgium), respectively. In each transfection experiment, to keep the amount of transfected DNA constant, the pCAT®-Control vector (GenBank: X65321.2; Promega Corp.) coding for a Chloramphenicol acetyltransferase was invovled as a neutral control expression vector. Two-Hybrid screening AD-vectors and DB-vectors were transformed into S. cerevisiae strains Y8800 (MATa) and Y8930 (MATα), respectively, using a one-step transformation protocol [77]. Yeast cells were plated onto synthetic dropout medium lacking tryptophan or leucine, respectively. Transformed yeasts were mated overnight at 30°C on solid medium containing yeast extract, peptone and dextrose (YEPD). Yeasts were then transferred on synthetic dropout medium plates lacking histidine, leucine and tryptophan to select diploids in which the GAL1-HIS3 reporter is activated. Control plates for autoactivation were composed of synthetic dropout medium containing cycloheximide (1 mg/L) and lacking histidine and leucine. Mating controls were plated on synthetic dropout medium lacking leucine and tryptophan. Cell culture and treatment Culture cells were maintained at 37°C, in a humidified atmosphere with 5% C02. HEK293T cell line was grown in Protein co-precipitation HEK293T cells were transiently transfected with vectors for FLAG-tagged-Hoxa2 variants and GST-tagged candidate interaction partners. Thirty-nine hours after transfection cells were lysed for 30 min at 4°C in ice-cold IPLS lysis buffer (0.5% NP-40, 20 mM Tris-HCl pH 7.5, 0.5 mM EDTA, 120 mM NaCl, 10% glycerol) including protease inhibitor cocktail from Roche (#11873580001, Roche). Cells lysates were centrifuged for 5 min at 16000 g at 4°C. Supernatants were recovered and samples were incubated with rotation overnight at 4 °C with glutathione-agarose beads (#G4510, Sigma) pre-washed three times with ice-cold IPLS lysis buffer. Beads were washed three times with ice-cold IPLS. Beads were supplemented with Laemmli loading buffer for SDS-PAGE (10% SDS, 30% glycerol, 350 mM Tris-Cl pH 6.8, 600 mM DTT, 0.1% bromophenol blue) and boiled 5 minutes at 95°C. Samples were centrifuged and loaded on denaturing SDS-PAGE gel for analysis by western blotting. As controls, in parallel to protein co-precipitation, expression of fusion proteins in the samples was confirmed by western blotting. Immunocytochemistry HEK293T cultured on glass cover slips were transiently transfected with FLAG-tagged Hoxa2 and GST-tagged RCHY1 expression vectors and treated with MG132 proteasome inhibitor 24h after transfection. After overnight treatment, cells were rinsed in PBS solution and fixed for 30 min with 4% formaldehyde in PBS. Cells were further blocked with 10% lowfat milk in TBS-0.1% Triton X100 solution for 45 min at room temperature, followed by over-night incubation in TBS-0.1% Triton X100 solution at 4°C, with mouse anti-FLAG antibody (M2) (#F1804, Sigma) and rabbit anti-GST (#G7781, Sigma) used at 1:500 and 1:50 dilution, respectively. Cells were rinsed three times for 30 min in TBS-0.1% Triton X100 solution and incubated for 45 min at room temperature with FITC conjugated anti-mouse (#sc-3699, Santa Cruz) and TRITC conjugated anti-rabbit (#sc-2367, Santa Cruz) at 1:100 dilution in TBS-0.1% Triton X100 solution. Cells were rinsed three times and glass cover slips were mounted in Vectashield®-DAPI medium (Vector laboratories). Slides were then analysed by confocal microscopy (LSM710, Zeiss, Jena, Germany). Metabolic recoding of epigenetics in cancer Dysregulation of metabolism allows tumor cells to generate needed building blocks as well as to modulate epigenetic marks to support cancer initiation and progression. Cancer-induced metabolic changes alter the epigenetic landscape, especially modifications on histones and DNA, thereby promoting malignant transformation, adaptation to inadequate nutrition, and metastasis. Recent advances in cancer metabolism shed light on how aberrations in metabolites and metabolic enzymes modify epigenetic programs. The metabolism-induced recoding of epigenetics in cancer depends strongly on nutrient availability for tumor cells. In this review, we focus on metabolic remodeling of epigenetics in cancer and examine potential mechanisms by which cancer cells integrate nutritional inputs into epigenetic modification. Background Dysregulated metabolism is one of the most prominent features of cancer. Since the postulation of aerobic glycolysis (Warburg effect) in the early 20th century [1], metabolic reprogramming in cancer has been the subject of extensive research [2]. Cellular metabolism is reprogrammed at multiple levels in cancer: genetic, epigenetic, transcriptional, posttranscriptional, translational control, and posttranslational [3][4][5][6][7][8][9][10]. Consequently, the expression of a wide range of metabolism-related proteins, such as metabolite transporters and metabolic enzymes, are dysregulated in cancer cells [11]. Metabolism is reprogrammed in cancer cells through the action of cell-intrinsic and -extrinsic factors. Alterations in oncogenes and tumor suppressor genes cooperatively remodel metabolic pathways to satisfy biosynthetic demands of cancer cells [12]. At the same time, microenvironmental factors modulate metabolic reprogramming; these factors include nutritional [13], inflammatory, and immune elements in malignant tissue [14]. For example, metabolic activity and nutritional status of cancer cells strongly influence epigenetics, especially modifications on histone and DNA [15]. The metabolic reprogramming interacts with epigenetic regulation and signal transduction to promote cancer cell survival and proliferation [16,17], and to influence a broad range of biological processes [18]. This review summarizes recent advances in our understanding of metabolic recoding of epigenetics in cancer, with particular emphasis on how cancer cells encode nutrient input into the epigenetic landscape. modifiers, leading to epigenetic remodeling. The interaction between cellular metabolism and epigenetics as well as the disease relevance of this interaction have recently been reviewed [15,17]. The focus of the present review is how cancer metabolism modulates DNA methylation, histone methylation, and histone acetylation, as well as their connection with nutrient availability. SAM, α-KG, oxygen and histone/DNA methylation Histones are methylated on lysine and arginine residues [26], and this methylation can repress or activate gene transcription [20]. Lysine methyltransferase (KMT) and arginine methyltransferase (PRMT) utilize S-adenosyl homocysteine (SAM) as the methyl donor in histone methylation (Fig. 2a). The reverse reaction of lysine demethylation is catalyzed by the amine oxidases lysine demethylases (LSD) 1 and 2 [27] in a reaction dependent on flavin adenine dinucleotide (FAD), as well as by an α-ketoglutarate (α-KG)-dependent dioxygenase, which produces succinate in an oxygen-dependent reaction [28] (Fig. 2a). Both α-KG and succinate are intermediates of the tricarboxylic acid (TCA) cycle, indicating a functional correlation between the TCA cycle and α-KG-dependent demethylation. The enzyme that demethylates histone arginine residues is being actively investigated [29,30]. The protein has been proposed to be an oxygen-and α-KG-dependent dioxygenase similar to that responsible for lysine demethylation [29]. In this case, too, demethylation is linked to oxygen levels and the TCA cycle (Fig. 2a). Metabolic intermediates participate as substrates or coenzymes in nearly all epigenetic coding processes. In cancer, metabolic dysregulation interacts with nutritional status to modulate epigenetic marks on histones and DNA. This nutritional status is defined largely as the availability of carbon sources. Nutrient availability affects epigenetic regulation in cancer Glucose availability is reflected in histone and DNA modification in cancer Glucose and glutamine are the major carbon sources of most mammalian cells, and glucose metabolism is closely related to histone acetylation and deacetylation. Glucose availability affects the intracellular pool of acetyl-CoA, a central metabolic intermediate that is also the acetyl donor in histone acetylation [33] (Fig. 1). Glucose is converted to acetyl-CoA by the pyruvate dehydrogenase complex (PDC), which produces acetyl-CoA from glucose-derived pyruvate; and by adenosine triphosphatecitrate lyase (ACLY), which generates acetyl-CoA from glucose-derived citrate. PDC and ACLY activity depend on glucose availability, which thereby influences histone acetylation and consequently modulates gene expression and cell cycle progression [34,35]. Dysregulation of ACLY and PDC contributes to metabolic reprogramming and promotes the development of multiple cancers, such as lung cancer [36]. At the same time, glucose metabolism maintains the NAD + /NADH ratio, and NAD + participates in SIRT-mediated histone deacetylation [37] ( Fig. 1). SIRT enzyme activity is altered in various malignancies [25,36,[38][39][40][41], and inhibiting SIRT6, a histone deacetylase that acts on acetylated H3K9 and H3K56, promotes tumorigenesis [42,43]. SIRT7, which deacetylates H3K18 and thereby represses transcription of target genes, is activated in cancer to stabilize cells in the transformed state [44][45][46]. Interestingly, nutrients appear to modulate SIRT activity. For example, long-chain fatty acids activate the deacetylase function of SIRT6, and this may affect histone acetylation [47,48]. Glucose catabolism affects histone acetylation as well as histone and DNA methylation, since glucose-derived α-KG serves as a substrate in the reactions catalyzed by histone demethylases and TET family DNA dioxygenases [49] (Fig. 2a, b). Glutamine metabolism modulates cancer epigenetics Glutamine metabolism also contributes to the production of acetyl-CoA and α-KG, and glutamine oxidation correlates with the cell state-specific epigenetic landscape. Naive embryonic stem cells efficiently take up both glutamine and glucose to maintain a high level of α-KG to promote histone and DNA demethylation, which in turn helps maintain pluripotency [49]. Inhibition of glutamine oxidation affects histone modifications including H4K16ac and H3K4me3 in breast cancer cell lines, altering the transcription of genes involved in apoptosis and metastasis [50]. Acetate and other carbon sources as epigenetic metabolites Cancer cells absorb acetate and incorporate it into histones [51]. Acetyl-CoA synthetases (ACSSs) convert acetate to acetyl-CoA, which in turn serves as a major carbon source in lower eukaryotes, but not mammals. However, glioma cells and hepatocellular cancer cells utilize acetate as an alternative carbon source to sustain acetyl-CoA production [52,53] (Fig. 1). This compensates for the hypoxic, nutrient-poor microenvironment of solid tumors. Mammalian cells express three ACSS isozymes (ACSS1-3). The contribution of ACSS isozymes to histone acetylation varies across different cancers [54][55][56]. ACSS is highly expressed in glioma and hepatocellular cancer, which correlates with histone hyperacetylation [54][55][56]. ACLY functions as a switch and controls carbon source preference of cancer cells [57]. Other carbon sources, such as fatty acids, also regulate epigenetic modifications [58] (Fig. 1). A high-fat diet reduces the acetyl-CoA level and decreases acetylation of H3K23 in white adipose tissue but not liver. This suggests that lipids may affect cancer risk via an epigenetic mechanism, since obesity predisposes to the development of multiple cancers [59]. One-carbon metabolism modifies chromatin methylation In one-carbon metabolism, the amino acids glycine and serine are converted via the folate and methionine cycles to nucleotide precursors and SAM. Multiple nutrients fuel one-carbon metabolism, including glucose, serine, glycine, and threonine [60] (Fig. 2a, b). High levels of the methyl donor SAM influence histone methylation [61], which may explain how high SAM levels prevent a b Fig. 2 Cancer cells coordinate nutrient status with the methylation of histone and DNA. Cancer cells alter methylation of histones (a) and DNA (b) in response to nutrient status. SAM S-adenosyl methionine, SAH S-adenyl homocysteine, KMT lysine methyltransferase, PRMT protein arginine methyltransferase, LSD lysine-specific demethylase, DNMT DNA methyltransferase, TCA tricarboxylic acid cycle, TET ten-eleven translocation methylcytosine dioxygenase malignant transformation [62]. Glucose availability is encoded in methylation of H3R17 by arginine methyltransferase CARM1 [63]. 2-hydroxyglutarate and oncometabolites In cancer, genetic alteration and microenvironment perturbation modify the catalytic properties of metabolic enzymes, reshaping epigenetics. Cancer-associated mutations in isocitrate dehydrogenase (IDH) 1 and 2 confer on the enzyme the ability to produce 2-hydroxyglutarate (2-HG), which is structurally analogous to α-KG [64] (Fig. 3). 2-HG competes with α-KG to bind to the catalytic pocket of several α-KG-dependent epigenetic enzymes, suppressing their catalytic activity and leading to genome-wide hypermethylation of histones and DNA [65,66]. The resulting aberrant gene expression promotes tumorigenesis [67,68]. The metabolic enzymes fumarate hydratase (FH) and succinate dehydrogenase (SDH) are also frequently mutated in certain cancers [69]. Loss-offunction mutations in FH and SDH lead to accumulation of fumarate and succinate, which act as competitive inhibitors of α-KG-dependent dioxygenase [70] (Fig. 3). The oncogenic effect of α-KG, fumarate, and succinate via epigenetic regulation has led them to be named oncometabolites [15]. 2-HG also accumulates in hypoxic cancer cells without IDH mutations, through a process mediated at least in part by the metabolic enzymes malate dehydrogenase (MDH) and lactate dehydrogenase (LDH). Hypoxia makes the tumor microenvironment acidic, which causes MDH and LDH to bind substrates promiscuously and generate 2-HG [71,72] (Fig. 3). Under these conditions, more 2-HG is produced by LDH than by MDH [73]. LDH may also modulate epigenetics in cancer cells independently of 2-HG, since tumor pH is highly heterogeneous and in fact only some cancer cell lines or tumor tissues reach

the pH of 6 needed to trigger promiscuous 2-HG production [74][75][76]. The in vivo significance of substrate promiscuity-induced 2-HG production remains to be explored. Other metabolites show oncogenic effects in certain tissues. For example, normal colonocytes utilize butyrate as a major carbon source. Glucose is used by a subtype of colon cancer cells as the carbon source, resulting in butyrate accumulation. Butyrate further inhibits HDAC to induce histone hyperacetylation and promote the proliferation of colon cancer cells [77] (Fig. 1). Conclusions Cellular metabolism is highly dynamic and compartmentalized. The accumulation of certain metabolites in cancer can target epigenetic enzymes to globally alter the epigenetic landscape. Evidence suggests that this alteration can be random. For example, cancer cells containing IDH mutations show highly variable DNA hypermethylation patterns, with effects on gene transcription difficult to predict [78]. In this model of metabolic recoding of cancer epigenetics (Fig. 4a), fluctuations in the level of a metabolite produce metabolic noise and randomly modify epigenetic marks to generate diverse clonal epigenetic landscapes. This provides an opportunity for clonal selection during tumor growth, metastasis, and relapse ( Fig. 4a). At the same time, recent studies have provided evidence supporting the idea that cancer-related metabolic changes lead to locus-specific recoding of epigenetic marks. Dose-responsive modulation of cancer epigenetics by metabolites 2-HG presumably inhibits all α-KG-dependent epigenetic enzymes, but its overall effects appear to depend strongly on its intracellular concentration. Cancer cells carrying IDH mutations, for example, vary significantly in 2-HG concentration [79], and this influences the resulting epigenetic recoding. Transient expression of mutant forms of IDH suppresses the H3K9 demethylase KDM4C more strongly than other demethylases [66]. In addition, α-KG-dependent dioxygenases show diverse half-maximal inhibitory concentrations (IC 50 ) of 2-HG [80]. These findings suggest that metabolic alterations in cancer cells reshape epigenetics in a manner dependent on metabolite dose (Fig. 4b). Histones are conjugated to a large number of metabolites [81]. It is thus reasonable to expect that fluctuations of metabolites can broadly impact the epigenetic landscape. Understanding metabolism-induced epigenetic alterations requires the development of an atlas of interactions between key metabolites and epigenetic enzymes in cancer cells. Sequence-specific recruitment of metabolic enzymes Precise recoding of epigenetic marks requires recognition of a specific genomic locus or DNA sequence. Metabolic enzymes that have translocated to the nucleus may recognize specific DNA sequences by binding to transcription factors (Fig. 4c). Some metabolic enzymes translocate to the nucleus in response to stress or physiological signals. For example, glucose deprivation causes cytosolic ACSS2 to relocate to the nucleus, where it binds to transcription factor EB (TFEB). When TFEB binds to the promoter regions of lysosomal and autophagy genes, it brings ACSS2 with it; the ACSS2 produces acetyl-CoA and increases histone H3 acetylation, modulating the expression of TFEB-regulated genes [82]. In a second example, glucose starvation enhances interaction between nuclear FH and ATF2. ATF2 recruits FH to its target genes, inhibiting H3K36me2 demethylation and increasing expression of those genes, ultimately arresting cell growth [83]. Other metabolic enzymes may also translocate to the nucleus and associate with transcription factors to mediate specific epigenetic remodeling. One hypothesis holds that the ability of nuclear ACSS2 to alter histone acetylation and of nuclear FH to alter methylation depend on high local concentrations of acetyl-CoA and fumarate, respectively, at the specific target DNA sequences [82,83]. Testing this hypothesis requires metabolite quantification in subcellular compartments, which remains a challenging task [84]. The engineering of artificial metabolite sensors may advance locus-specific and real-time monitoring of epigenetic metabolites [85]. Studies are also needed to explore the possibility that nuclear metabolic enzymes modify epigenetic marks independently of their catalytic activity. Targeting of epigenetic enzymes by nutritional signals Nutrient sensing and signaling is a key regulator of epigenetic machinery in cancer. During glucose shortage, the energy sensor AMPK activates arginine methyltransferase CARM1 and mediates histone H3 hypermethylation (H3R17me2), leading to enhanced autophagy [63]. In addition, O-GlcNAc transferase (OGT) signals glucose availability to TET3 and inhibits TET3 by both decreasing its dioxygenase activity and promoting its nuclear export [86]. These observations strongly suggest that nutrient signaling directly targets epigenetic enzymes to control epigenetic modifications (Fig. 4d). The nutritional status of cancer cells is highly dynamic during cancer development. How cancer cells coordinate nutrient status with epigenetic phenomena during cancer progression remains an open question. Concluding remarks Our understanding of cancer metabolism has increased tremendously in the last decade. What were once considered bystander cells in the tumor microenvironment-such as cancer-associated fibroblasts [87], immune cells, and inflammatory cells [88,89]-are now recognized as contributors to metabolic remodeling of cancer [90]. Dose-dependent effect of metabolites on epigenetic enzymes. Higher accumulation of a specific metabolite affects more epigenetic targets. Half-maximal inhibitory concentrations (IC 50 ) of different target epigenetic enzymes are indicated as triangles. Different colors of triangles represent different epigenetic enzymes. c Metabolic enzymes translocate to the nucleus, where they bind to transcription factors that carry the enzymes to specific target sequences in the genome. d Nutrient sensing and signaling modulate the epigenetic machinery microenvironment [91], while pancreatic cancer cells depend on alanine secreted by stroma-associated pancreatic stellate cells [92]. Metabolite transport within tumor tissue and crosstalk between cancer cells and "bystander" cells cooperatively remodel cancer metabolism, suggesting an intricate and complicated regulatory network in the tumor microenvironment. Metabolic remodeling has also been implicated in a variety of human diseases other than cancer [17,93]. Cellular metabolism is closely related to stem cell homeostasis and differentiation [94]. Elucidating the connection between metabolism and epigenetics would provide mechanistic insights into these diseases and offer potential therapeutic opportunities for translational investigations. Authors' contributions YPW and QYL formulated the idea for this review, which they co-wrote. Both authors read and approved the final manuscript. Evaluation of the effect of oral appliance treatment on upper-airway ventilation conditions in obstructive sleep apnea using computational fluid dynamics ABSTRACT Objective: To evaluate the effect of oral appliance (OA) treatment on upper-airway ventilation conditions in patients with obstructive sleep apnea (OSA) using computational fluid dynamics (CFD). Methods: Fifteen patients received OA treatment and underwent polysomnography (PSG) and computed tomography (CT). CT data were used to reconstruct three-dimensional models of nasal and pharyngeal airways. Airflow velocity and airway pressure measurements at inspiration were simulated using CFD. Results: The apnea–hypopnea index (AHI) improved from 23.1 to 10.1 events/h after OA treatment. On CFD analysis, airflow velocity decreased at the retropalatal and epiglottis-tip levels, while airway pressure decreased at the retropalatal, uvular- and epiglottis-tip levels. The AHI of patients with OSA before OA treatment was correlated with airway pressure at the epiglottis-tip level. Discussion: Treatment with OA improved the ventilation conditions of the pharyngeal airway and AHI. Results of CFD analysis of airway pressure and airflow velocity helped determine the severity and ventilatory impairment site of OSA, respectively. Introduction Obstructive sleep apnea (OSA) is an obstructive respiratory disorder that occurs repeatedly during sleep. Consequently, sleep is fragmented, which results in the shortage of sleep; thus, patients experience somnolence during the day, leading to a decrease in activity. Moreover, when left untreated, OSA can cause long-term complications, such as cardiac, circulatory, and endocrine diseases, and mental illness. Several adverse effects of OSA on vital prognosis have been reported [1][2][3]. Currently, mainstream therapies for OSA include weight loss for obese patients and conservative treatment using continuous positive airway pressure (CPAP) and oral appliances (OAs) [4][5][6]. Although CPAP treatment is the gold standard, there are cases in which the CPAP apparatus cannot be employed due to issues with its operability or portability or a sense of continuous incongruity with the mask [7]. Guidelines issued by the American Academy of Sleep Medicine (AASM) state that OA treatment is indicated in patients with mild-to-moderate OSA and in those for whom CPAP is not indicated [8]. However, severe OSA has been shown to respond to OA treatment. Johal et al. [9] reported that six of eight subjects with severe OSA were successfully treated with OA treatment. Therefore, the indication for OA treatment cannot be determined solely on the basis of OSA severity. To determine whether OA treatment is indicated for a patient, it is necessary to evaluate and consider not only the apnea-hypopnea index (AHI) but also upper-airway ventilation conditions. Previously, cephalometry [10], computed tomography (CT) [11], and endoscopy [12] were used for the evaluation of upper-airway ventilation conditions during OA treatment. However, because the upper airway is complicated and extends from the nose to the hypopharynx, the positive effects of OA treatment on ventilation conditions cannot be determined solely by morphologic observation. Recent OSA-airflow studies have employed computational fluid dynamics (CFD) for analysis [13][14][15][16][17]. In two of those studies, the pharyngeal airway ventilation conditions were changed by use of rapid maxillary expansion and Herbst appliance using CFD [15,16]. Additionally, Mihaescu et al. [14] evaluated changes in the ventilation conditions of the upper airway by maxillomandibular advancement surgery using CFD. The authors of the present study believe that CFD is very effective for the evaluation of upper-airway ventilation conditions, as it can evaluate the flow of air in a manner similar to that during actual breathing, even in cases of upper airways with complicated morphologies. The large negative pressure measured in the pharyngeal airway by CFD may predict pharyngeal-airway collapse during sleep [18,19]. Moreover, functional evaluation of the upper airway using CFD might provide new findings that are not available from morphologic evaluation alone. This retrospective study was conducted for the purpose of evaluating the effect of OA treatment on upperairway ventilation conditions in OSA using CFD. Subjects This retrospective study included 15 patients (13 men and 2 women) with OSA who first visited Yamaguchi University Hospital in Ube, Japan between June 2009 and March 2012. The mean age and body mass index (BMI) of the study group were 51.3 years and 23.9 kg/m 2 , respectively ( Table 1). The mean AHI, apnea index (AI), hypopnea index (HI), and lowest oxygen saturation level (SpO 2 ) were 23.1 events/h, 10.5 events/h, 12.6 events/h, and 82.1%, respectively. Those with mild to moderate OSA, determined by polysomnographic (PSG) examinations, received OA treatment. In addition, those with severe OSA who dropped out of CPAP treatment received OA treatment [8]. Patients were administered OA treatment with one of two types of mandibularrepositioning devices: seven patients with a history of temporomandibular disorder (TMD) symptoms, such as clicks, received duoblock-type OAs [20][21][22], while eight patients without the TMD symptoms received monoblock-type OAs [21,23,24]. For both OA types, the mandibular protruded positions were titrated. Initially, mandible advancement was set at 50% of the maximum mandibular protruded position. OAs were incrementally titrated according to either a maximal comfortable protruded position of the mandible or a resolution of snoring and daytime symptoms [25]. Isacsson et al. [26] reported that there was no significant difference in the treatment effects between the duoblock-type and monoblock-type devices. In this study, there was no significant difference in the amount of mandibular advancement set with either device (monoblock vs. duoblock). Additionally, there was no significant difference in the PSG data (AHI, AI, HI, or SpO2) measured before and after the treatment. Therefore, the present study analyzed the data of the monoblock and duoblock OA groups together. This study was performed under the approval of Kagoshima University (Kagoshima, Japan; No. 180073 (657) Epi-ver.1) and Yamaguchi University (Yamaguchi, Japan; No. H28-138). Because of the retrospective nature of this study, the requirement for informed consent was waived, and an exemption was granted in writing by the institutional review board. Polysomnographic and CT examinations All patients underwent PSG (Alice 5 Diagnostic Sleep System, Philips Respironics, Best, The Netherlands) and CT during the initial consultation and after confirming that there was no problem with the use of OAs after the titration. After the symptom improvement PSG and CT examinations were performed at a mean of six months following the start of OA treatment, with the OA in place. The AHI, AI, HI, and lowest SpO 2 level were measured by PSG. Apnea was defined as complete cessation of airflow for 10 s, and hypopnea was defined as a 50% reduction in oronasal airflow lasting 10 s with at least 3% desaturation. The AHI was calculated as the number of apnea and hypopnea events per hour of sleep [8]. During the CT examination, each patient was asked not to move his or her head or swallow at the end of

expiration. Multi-slice helical CT (SOMATOM Definition, Siemens AG, Erlangen, Germany) images of the upper airway were acquired in supine position in order to reproduce sleep-related breathing conditions. The slice thickness was 0.6 mm. Imaging data were transferred directly to a personal computer and stored in the Digital Imaging and Communications in Medicine format. Before the CT exam, the patients were fully informed of the purpose and risks of the procedure. Evaluation of upper airway ventilation conditions Three-dimensional (3D) reconstructions of the upper airway were generated from the CT data using volume-rendering software (Intage Volume Editor; Cybernet, Tokyo, Japan). The airway was segmented primarily on the basis of image intensity with the threshold set midway between the soft tissue and clear airway value. As a result, the threshold was shown from −550 to 1024 Hounsfield Units [27]. Subsequently, the 3D models were converted into smoothed models using mesh-morphing software (DEP Mesh Works/Morpher; IDAJ Co., Kobe, Japan), without losing the patient-specific shape of the upper airway. The ventilation conditions of these upperairway models were then evaluated by CFD. The models were exported to a fluid-dynamics program (Phoenics; CHAM-Japan, Tokyo, Japan) in stereolithographic format. The fluid was assumed to be Newtonian, homogeneous, and incompressible. Elliptic-staggered equations and the continuity equation were used for analysis. The CFD of the upper-airway models were analyzed at a volumetric flow rate of 500 cm 3 /s and a no-slip condition at the wall surface [28]. Mean values were calculated after 500 iterations. Convergence was judged by monitoring the magnitudes of absolute residual sources of mass and momentum normalized to their respective inlet fluxes. Iteration was continued until all residuals fell below 0.2% [15,16]. The simulation estimated the airway pressure and airflow velocity at the hard-palate, retropalatal, uvular-tip, and epiglottis-tip levels ( Figure 1). Statistical analysis In this study, no significant differences in variables were found in relation to sex and age. Therefore, statistical analysis was performed using a pooled sample. For all variables, the paired t-test was performed to determine the significance of the treatment-associated changes detected after PSG. Whenever a variable exhibited nonnormal distribution of data or differing variance, the significance of treatment-associated changes after CFD was determined using the nonparametric Wilcoxon rank test. Intersite differences were determined using the Friedman test with Bonferroni correction. Spearman's correlation coefficients were calculated to evaluate the relationships between PSG data and maximal negative pressure within the four parts of the pharyngeal airway. Statistical significance in all tests was set at p < 0.05. A power analysis was performed to calculate the β error (1 -β error = 0.80; α = 0.05; two-tailed test); the target sample size was 13 subjects. Therefore, the present sample size was sufficient for this analysis. All measurements were repeated after one week by the same investigator (HS), and method errors were calculated using Dahlberg's formula [29]. The method errors of maximal airway pressure and airflow velocity were 1.712 Pa and 0.062 m/s, respectively. On the basis of the results of repeated analyses, these method errors were considered negligible. Results of CFD analysis Before OA treatment, airflow velocities at the retropalatal and uvular-tip levels were faster than those at the hard-palate level (Table 3), while negative-pressure levels at the uvular-and epiglottis-tip levels were higher than those at the hard-palate level ( Table 4). After OA treatment, airflow velocities at the retropalatal and epiglottis-tip levels significantly decreased from 9.0 to 5.3 m/s (p = 0.006) and 6.8 to 4.8 m/s (p = 0.024), respectively (Table 3). Moreover, negativepressure levels at the retropalatal, uvular-tip, and epiglottis-tip levels significantly decreased from −130. (Table 4). Before OA treatment, AHI was correlated with negative pressure at the epiglottis-tip level (r s = −0.521; p = 0.046; Table 5). However, after OA treatment, AHI was not correlated with negative pressure at any of the levels (Table 6). In addition, post-treatment change in AHI was correlated with that in negative pressure at the uvular-(r s = −0.582; p = 0.023) and epiglottis-tip (r s = −0.536; p = 0.040; Table 7) levels. Discussion The present study investigated changes in the airway ventilation conditions after OA treatment in patients with OSA by means of CFD analysis of upper-airway models extending from the nasal cavity to the hypopharynx. Relationship between AHI and airway pressure in CFD models In a previous study that performed CFD analysis of upper-airway ventilation conditions at inspiration in obese children with adenoidal hypertrophy, Wootton et al. [18] reported that negative pressure of the pharyngeal airway was closely associated with the severity of OSA. Therefore, the present study examined the relationship between airway pressure and PSG data in patients with OSA before OA treatment and found a correlation between pretreatment negative airway pressure at the epiglottis-tip level and AHI. Downing and Ku [30] reported that large negative pressure in the pharyngeal airway at inspiration induces pharyngeal-airway collapse. In the present study, the pretreatment negative pressure at the epiglottis-tip level was large and correlated with pretreatment AHI. The epiglottis-tip level is the anatomical site that can easily induce glossoptosis. Therefore, the present study concluded that negative pressure at the epiglottis-tip level is correlated with pharyngealairway collapse. Additionally, it is easy to physically transform the airway at the retropalatal and uvular-tip levels. It is, therefore, believed that pharyngeal-airway collapse is easily induced at these levels. In the present study, although negative-pressure levels at the retropalatal (r s = −0.450; p = 0.092) and uvular-tip (r s = −0.496; p = 0.060) levels were not significantly different, they showed a weak correlation with AHI; they showed the same tendency at the epiglottis-tip level. It is possible that these results might have been influenced by the small sample size. Nevertheless, these results suggest that CFD analysis of negative-airway pressure at inspiration is an effective method for evaluating the severity of OSA, which the authors believe to be a novel finding because it cannot be obtained solely by morphological observation. Airflow velocity in CFD models In a previous study, Downing and Ku [30] reported that airflow velocities obtained from CFD models are affected by the size and shape of the airway and that a narrowed airway is subject to higher airflow velocities than an unobstructed one. In other words, in a CFD model, the site with the higher airflow velocity indicates the region of ventilatory impairment. In the present study, airflow velocities at the retropalatal and uvular-tip levels in OSA patients before OA treatment were faster than those at the hardpalate level. The authors believe that, in these patients, airways at the retropalatal and uvular-tip levels were narrow, which caused ventilatory impairment. In many studies, endoscopy findings have revealed that patients with OSA exhibit obstruction in the velopharynx [31,32]. The retropalatal and uvular-tip levels defined in the present study correspond to the velopharynx, and the results observed in this study are similar to those reported in previous studies. These findings indicate that airflow velocities determined by CFD analysis might be effective for detecting sites of ventilatory impairment in the airway. CFD findings after OA treatment The present study evaluated the changes in upperairway ventilation conditions after OA treatment by performing CFD analysis. After OA treatment, airflow velocities at the retropalatal and epiglottis-tip levels, as well as negative-pressure levels at the retropalatal, uvular-tip, and epiglottis-tip levels, decreased significantly relative to pretreatment levels. Chan et al. [33] evaluated morphological changes of the pharyngeal airway after OA treatment on the basis of magnetic resonance imaging findings and found that the volume of the velopharynx and hypopharynx had increased after treatment. Additionally, the present findings indicated that OA treatment might help enlarge the pharyngeal airway at the retropalatal and epiglottis-tip levels, leading to improvement of ventilation conditions in the pharyngeal airway. Changes in CFD and PSG data after OA treatment In the present study, the post-treatment change in AHI was correlated with that in negative pressure at the uvular-and epiglottis-tip levels. This suggests that it is important to improve the negative-pressure levels at the uvular-and epiglottis-tip levels to achieve a decrease in AHI. In an endoscopy study, Sasao et al. [12] evaluated morphological changes in the pharyngeal airway when the lower jaw was moved forward. The authors found that the velopharynx enlarges with OA treatment, which easily allows improvement of AHI. Expansion of the velopharynx, leading to ventilatory impairment, has an influence on downstream airflow. It is possible that, in the present study, the velopharynx was enlarged because of OA treatment, leading to improvement in ventilatory impairment. This, in turn, might have helped relieve the negative pressure at the uvular-and epiglottis-tip levels, causing a decrease in AHI. Validity of CFD models In a study by Zhao et al. [34], pharyngeal-airway pressure before OA treatment in patients with OSA was approximately −50 Pa, which is lower than the pressure level observed in the present study. However, the flow rate reported in the previous study was one-third of the value observed in the present study. Therefore, the authors believe that the present results show a similar tendency as those reported by Zhao et al. and that the CFD methodology is accurate in the present study. Case presentation Although there have been several reports on the eligibility criteria for OA treatment and the effect of OAs on OSA [8,9,12,35], there are few reports on the reason for failure of OA treatment in some instances. This section depicts the results of CFD analysis of patients who were responsive or nonresponsive to OA treatment. One of the responsive patients exhibited severe OSA and a pretreatment AHI of 50.3 events/h. However, after OA treatment, the AHI had decreased to 15.7 events/h, which indicated a 68.8% improvement (Figure 2). The results of CFD analysis revealed that OA treatment had led to improvement of airflow velocity and negative pressure from the retropalatal level to the epiglottis-tip level. Although this patient did not meet the AASM eligibility criteria for OSA severity [8], the authors believe that improvement of ventilatory impairment with OA treatment led to a decrease in AHI in this case. In a representative nonresponsive case, in which the AHI did not improve upon OA treatment, a patient exhibited a pretreatment AHI of 20.1 events/h and moderate OSA in accordance with the AASM guidelines; this was a case in which OA treatment was indicated [8]. The pretreatment airflow velocity and negative pressure at the epiglottis-tip level were high. After treatment, the AHI in this case had increased to 29.0 events/h (Figure 3a,b), and negative-pressure levels had increased in the entire upper airway. In the present study, these findings were confirmed by analyzing a sagittal section of the left side of the nose (Figure 3c), which revealed a site in the nasal airway that exhibited relatively high airflow velocity; the negative-pressure level downstream from the site was also high. In this case, although OA treatment helped improve the airflow velocity at the epiglottis-tip level, the AHI was thought to have increased because of a problem in the nose that had occurred during treatment. The findings of these two cases suggest that CFD analysis of not only the pharyngeal airway but also the upper airway (including the nasal airway) is effective for detecting sites of ventilatory impairment and might contribute to improvement of treatment outcomes in OSA. Limitations The upper-airway model constructed in this study used data generated in the wake state (not the sleep state); therefore, it did not include an adjustment for transformation during breathing. However, pharyngeal-airway pressure at inspiration and in deep stages of sleep with complete muscle atonia and airway collapse influence airway transformation during sleep. However, the authors believe that the results obtained by analysis of data generated in an awakened state may be useful because the present data revealed a correlation between pharyngeal airway pressure and sleep parameters. With regard to OA treatment, because of differences in the clinical states of individuals, the present study did not administer a uniform type of OA or achieve a uniform change in mandibular position. This was a preliminary study, and the authors intend to perform detailed and individual analyses in a larger sample in the future. Furthermore, a randomized sampling approach was not used, which may have introduced a bias based on the available patients; this should be considered when . A case of mild-to-moderate OSA in which the symptoms did not improve with OA

treatment. (a) Before treatment -The AHI was 20.1 events/h. Left: The airflow velocity was extremely high at the epiglottis-tip level (large red arrow). Right: The negative pressure was extremely high at the epiglottis-tip level (large red arrow). (b) After treatment -The AHI had worsened to 29.0 events/ h. Left: The airflow velocity at the epiglottis-tip level had decreased after OA treatment (small yellow arrow). Right: However, the negative pharyngeal-airway pressure had become even more severe (large red arrow). (c) After treatment -Left: Frontal view of the upper airway; the red line cuts off the left part of the nose. Right: Sagittal view of a plane on the left side of the nose. The airflow velocity at the nose was extremely high, which indicated the site of ventilatory impairment (large red arrow). In this case, OSA was thought to be caused by nasal-airway obstruction. OSA: obstructive sleep apnea; OA: oral appliance; AHI: apnea-hypopnea index. interpreting the findings of this study. A future study should include a randomized approach, in order to reduce the risk of bias during patient sampling. Conclusion The present study examined the changes in upper-airway ventilation conditions after OA treatment using CFD. The results demonstrated that OA treatment improved the ventilation conditions of the pharyngeal airway and significantly improved the AHI. Additionally, airway pressure and airflow velocity data obtained from CFD models demonstrated the severity of OSA and the site of ventilatory impairment in OSA, respectively. Stress Level Comparison of Medical and Non-medical Students: A Cross Sectional Study done at Various Professional Colleges in Karachi, Pakistan Objective: To compare the stress levels of medical students with that of other professional colleges. Background: Stress is known to affect learning abilities and also be a risk factor for various health and psychological difficulties. Through earlier studies, stress levels of medical students have established to be high during their academic life. In Pakistan, local epidemiological data about psychological morbidity among medical undergraduate students is infrequent. An extensive electronic Internet-based search failed to locate any study which shows a comparison of stress between medical students and the students of other professions in Karachi, which is the objective of our study. Methods: The study was conducted at various professional colleges all over Karachi. A sample of 600 students, 50 from each of the 12 selected colleges was taken. A standardized stress questionnaire of the International Stress Management Association (UK) was used to assess the stress levels which categorized the level of stress into mild, moderate and severe. Results: Stress levels were found to be higher in medical students, and this stress was mostly attributed to studies according to majority of the medical students (75.6%), where as calculated stress levels were also higher in medical students (54.6%) Conclusion: Stress levels of medical students were found to be suggestively higher than those of non-medical professional students. Thus, medical students should be provided with appropriate counseling and stress relieving activities to prevent the long term antagonistic effects of elevated stress levels on the physical and mental health of future doctors. Introduction Stress can be defined as ''a state of mental or emotional strain or suspense'' and also as ''a number of normal reactions of the body (mental, emotional, and physiological) designed for selfpreservation'' (Princeton University, 2001). Despite its diffuse perception, most of the well-known definitions emphasize stress as ''any factor that threatens the health of an individual or has an adverse effect on the functioning of the body'' (Oxford Medical Publications, 1985). It has been proven through various researches that medical students experience elevated stress levels throughout their medical school life [1,2]. The personal and social expense they have to make in order to maintain respectable academic results, in a highly competitive environment, puts them under a lot of stress [3]. Stress can lead to interruptions in both physical and mental health. It may lead to the development of depression and anxiety coping with the help of drugs, analgesics; alcohol, smoking and eating actually are counterproductive and may calculated out of a total of 25 and categorized into low (score=6 or less), moderate (score=7- 14) and high (score=more than 14). The low, moderate and high scores were reflecting the level of stress found out to be present in these students. Additionally, 2 more questions were added to check for perceived stress levels and stress attributable to studies. Study subjects Twelve professional colleges of Karachi were selected at random, 3 from each of the four following fields. 50 students were selected, by random sampling from each of the 12 colleges, making a total sample size of 600. The data collection was carried out over a period of 2 months (April-May). Inclusion and exclusion criteria: Students of 2 nd and 3 rd years were selected. All students other than the ones in 2 nd and 3 rd year were excluded as stress levels are perceived to be high at the beginning and towards the end of the professional education. Questionnaires were to be filled out by the students in the mid of the term ensuring that they weren't about to give any examinations as stress levels would be invariably high during that period. Only students from the private colleges were selected specifically; excluding all Government colleges. Collection of data: We personally visited each of the colleges. The students who fitted selection criteria were selected at random. The students were allowed to respond in their own time and privacy. The participation was entirely voluntary. Results 50 students from each of the 12 colleges mentioned above filled out the questionnaires. The stress levels were calculated out of a total of 25, and categorized into low, moderate and high. Out of these, 54.6% medical students have stress levels in the 'high' range compared to the 20.6% of engineering, 20.6% of arts and 32% of commerce students in the same category. Most of the non-medical students fall into the 'moderate' stress levels range with the highest being that of the engineering students 74.6%, then arts students 68.6% and finally commerce students 68% (Table 1a). In 'perceived stress levels' (percentage of students who considered their stress levels to be higher than 6 on any given day) 54.3% of medical students responded with a yes while 15.3% of engineering students, 32% of commerce and 36.6% of arts students had the same response. About 75.6% medical students worsen the stress. While some students see the pressure as challenge to work harder, others find it hard to cope with the stress and lag behind. Stressors (triggers) are anything that tosses the mind-bodyspirit connection out of equilibrium (homeostasis). When your mind/brain construes stressors (triggers) to be a threat to internal balance and synchronization, then a protective "good stress" response is generated in order to restore homeostasis. Allostasis is the process of attaining constancy (homeostasis) through change, which functions by way of the mind-body-spirit communication systems. Allostasis is central for your health and survival. Researchers from Heptares Therapeutics, a drug company, have exposed the 3-D structure of a protein receptor that mediates our response to stress, CRF1, a molecule on the outside of cells on the pituitary gland, releases CRF, hormones involved in regulating our stress response that over time underwrite to anxiety and depression. A vital aspect of the discovery is that the receptor has a small binding pocket positioned in a much dissimilar position than other G-protein-Coupled Receptors, (GPCRs). Knowing the structure of Class B GPCRs like CRF1 could possibly help researchers develop drugs that better target receptors within the same family, Heptares claims. Scientists could just enterprise a drug that pops right into that pocket. Additionally, research suggests that stress can block chemical responses in the brain that are necessary for learning. Stress can disrupt learning and memory progress (Long-Term Potentiation (LTP)) as it forces the brain to revert to more primitive survival needs. Although it is known that long-term or chronic stress can affect the brain's learning and memory region, a new finding discovers short-term stress, lasting as little as a few hours, can also impair brain-cell communication in these critical areas [4]. Our research is based on our hypothesis that medical students are subjected to even more stress than students of other professions, which in itself is immensely counter productive as no other profession demands better mental and social well being than the medical profession, as there is no aspect of the job that does not require public interaction or the ability to make quick, responsible decisions. Instrument The instrument used was the Standardized Stress Questionnaire formulated by the International Stress Management Association UK. The questionnaire was slightly modified to better adapt to our topic of study. It consisted of 25 questions, along the lines of 'Do you study till late in the night?' and 'Do you think there are too many deadlines in your study life that are difficult to meet' to assess the academic stress levels of individual students. The questions were to be answered in either yes or no, which was designed so to be very easy for the student rather than scoring which doesn't give a clear answer. For each yes, there was one mark while each no was valued to be zero. The scores were have attributed their high stress levels to studies (p-value=0.000), which is more than twice that of students of other professions as compared to engineering 20%, commerce 38% and arts 36.6%. A number of variables showed results that could explain this difference in statistics such as: around 81.3% medical students complained that they had too much to study followed by commerce 62.6%, engineering 60% and arts 48% as shown in Table 1b. Approximately 76% medical students thought that they had too many deadlines in their life followed by commerce 58%, engineering 48% and arts 46% as shown in Table 1b. Discussion Studies have shown that medical students experience a high level of stress during their undergraduate course [5-9]. High level of stress may have a negative effect on cognitive functioning and learning abilities of students in the medical school [10]. The estimated prevalence of emotional disturbance found in different studies on medical students was higher than that in the general population. In three British universities, the prevalence of stress among medical students was 31.2% [11], and 41.9% in a Malaysian medical school [12] and 61.4% in a Thai medical school [13]. In a Swedish study, the prevalence of depressive symptoms among medical students was 12.9%, and 2.7% of students had made suicidal attempts [10]. Thus, the stress levels of medical students have been found to be significantly higher than those of the other 3 professional colleges (Figure 1). The factors that could be attributed to this difference are: The vastness of the course that has to be covered in a relatively short amount of time, too many deadlines that are difficult to meet (p value=0.00) and the inability to relax due to feelings of guilt while relaxing (p value=0.00) ( Table 2). When asked if the stress was due to studies, 75.6% of medical students responded with a yes whereas 20% of engineering, 38% of arts and 35.6% of commerce students answered yes ( Figure 2a and 2b). The perceived stress levels were high in 54.3% of medical students as compared to the 54.6% who had high levels of calculated stress (true positives). This implies that a great percentage of students are unaware of their elevated stress levels, which could be potentially dangerous because these students are unlikely to seek counseling or help to reduce their stress levels. The negative effects of long and tiring medical education on the psychological status of students have been shown in several studies. Results of a study in the UK showed that one-third of psychologically-ill students did not graduate from the college [14]. Our study showed three aspects of the effects of high stress levels on an individual namely: physical, psychological and professional. The physical effects can be seen in the form of loss of appetite (p value=0.000), constant fatigue (p value=0.000), increase in muscular aches (p value=0.006); the psychological problems in the form of critical personality (p value=0.046), impatience (p value =0.000) and inability to relax (p value=0.000); and professional in the form of being constantly preoccupied with own thoughts(p value=0.000), clouded judgment and inability to concentrate (p value=0.001) Table 2. Therefore, the short term

effects would be that the lack of physical fitness would reduce the students' ability to learn and perform well in their exams as they are unlikely to be able to focus on lectures during classes. The long term effects of these would be increased likelihood of developing chronic diseases such as ulcers, hypertension and diabetes, psychological abnormalities such as anxiety and depression and poor professional abilities due to lack of knowledge, poor judgment and inability to make quick and sensible decisions. Therefore, with early identification and effective psychological services, possible future illnesses may be prevented. It is very important to target stress-prevention strategies at students who have any level of psychological stress to prevent the development of more serious conditions relating to stress. Wellness and mental health programs are also needed to help students make smooth transition between different learning environments with changing learning demands and a growing burden on their mental and physical capacity. Medical schools in the United States and Canada have initiated health-promotion programs and have reported positive results in reducing the negative effects of stress upon health and academic performance of medical students [15][16][17]. A similar approach to reduce the level of stress could be used for the medical students Karachi. On the other hand, a minimal amount of stress is necessary to add spice to one's life and to achieve optimal performance at examinations. An element of stress is involved with growth and is essential for sound personal functioning. Students should be encouraged to participate in extra curricular activities by allocating proper time and giving incentives such as extra points in final examination or awards and cash prices to distract the students from the burden of their stressful academic life. Limitations This cross-sectional study was based on self-reported information provided by students. Therefore, there is some potential for reporting bias which may have occurred because of the respondents' interpretation of the questions or desire to report their emotions in a certain way or simply because of inaccuracies of responses. Recommendations Another longitudinal study should be carried out with a cohort of students to investigate the levels of stress among students in all the five years of education. Conclusion The findings of the study recommend that the level of stress was higher amongst medical students as compared to students Figure 1 Stress levels of the students (Arts and Engineering). of other profession. These augmented stress levels in turn are responsible for symptoms such as decreased appetite, impaired attention and clouded judgment. These findings also suggest that special care must be taken to find out the obvious psychiatric problems among them. The major finding of high stress in medical students points to the need for establishing counselling and preventive mental health services as an integral part of routine clinical services being provided to the medical students and initiatives must also be taken by the governing body to bring about a change in the curriculum which may help in decreasing stress due to studies. Students should also be encouraged to indulge themselves in various extracurricular activities which may have a refreshing effect on their minds. Agreement of wavefront-based refraction, dry and cycloplegic autorefraction with subjective refraction Purpose To evaluate the agreement of dry, and cycloplegic autorefraction and wavefront-based refraction with subjective refraction. Method 83 subjects aged 19–57 years were included in this cross-sectional study. Refractive status was determined using four methods including subjective refraction, wavefront-based refraction, dry and cycloplegic autorefraction. Refractive data were recorded as sphere, cylinder and spherical equivalent (SE). Power vector components were used to compare the astigmatism obtained using the different methods of refraction. Results The more negative spherical, cylindrical and SE components were obtained using dry autorefraction, wavefront-based refraction and dry autorefraction, respectively. The less negative spherical, cylindrical and SE components were obtained using cycloplegic autorefraction, subjective refraction and cycloplegic autorefraction, respectively. Considering the spherical component, there was a statistically significant hyperopic shift (0.12 ± 0.29 D, p = 0.001) with cycloplegic autorefraction and a significant myopic shift (−0.17 ± 0.32 D, p < 0.001) with dry autorefraction compared to subjective refraction, while the difference between wavefront-based and subjective refraction was not significant statistically (p = 0.145). The calculated cylindrical component using subjective refraction showed statistically significant difference with dry auto-refraction (p < 0.001), cycloplegic auto-refraction (p = 0.041) and wavefront refraction (p < 0.001). The highest correlation with subjective refraction in sphere, cylinder and SE was observed for cycloplegic auto-refraction (rs = 0.967), dry auto-refraction (rs = 0.983) and cycloplegic auto-refraction (rs = 0.982), respectively. Conclusions As subjective refraction is gold standard in our study, sphere in cycloplegic auto-refraction and astigmatism in dry auto-refraction showed better agreement and correlation. Introduction Various techniques have been developed to assess the refractive status of the eye. Traditionally, subjective manifest refraction has been known as a gold standard for refraction and spectacle prescription, although it is a timeconsuming procedure. 1 Objective autorefracometers and, more recently, wavefront-based autorefractometers have achieved their clinical popularity because of their acceptable repeatability, accuracy, time-saving capability and ease of use. 2---6 The introduction of automated refraction in the 1970s has created a promising revolution in eye examination. 7 Autorefraction is considered as a clinically valuable starting point for subjective refraction. 8---10 Over the past few years, the wavefront analysis systems was considered as a turning point for clinical optics because these systems provide a new way to examine the refractive errors in details, including lower and higher orders aberrations. 11 The difference between classical refraction and wavefront sensing is analogous to the difference between keratometry and corneal topography. Subjective refraction and autorefraction both have a limited ability to separate between low and higher orders aberrations such as coma, trefoil and spherical aberration that can adversely affect the quality of retinal images and ultimately the visual quality. 12 Wavefront sensing devices supply both types of aberrations, lower (spherical defocus and astigmatism) and higher orders aberrations.The wavefront-based refraction is based on fitting a reference (ideal) wavefront on the actual wavefront. 13 The subjective refraction is considered as a standard procedure in assessing refractive errors to consider the need to modify the present correction. Sometimes, subjective refraction is not possible, as in young children or uncooperative subjects, and the findings of other techniques must be used. Therefore, the present study was designed to evaluate the agreement of the results obtained using wavefront-based refraction, dry and cycloplegic autorefraction and subjective refraction in adults. Methods This cross-sectional study included 83 eyes of 83 subjects aged 19---57 years who met the inclusion criteria. All procedures of this study followed the principles of the Declaration of Helsinki, and an informed consent was obtained from each subject after explaining the goals of the study. The study protocol was approved by the Ethics Committee of Shiraz University of Medical Sciences. (Code No.: 16930) Exclusion criteria were corrected distance visual acuity of less than 20/25, history of corneal refractive surgery, previous eye trauma, corneal scar and corneal irregularity secondary to ectasia or dry eye, cataract and history of long-term contact lens wearing. Refractive status was determined for each subject using four methods, including subjective refraction, wavefront-based refraction, and dry and cycloplegic autorefraction. All techniques were performed by the same experienced and qualified optometrist to avoid any bias during a single visit. Subjective refraction was done using maximum plus acceptance for all subjects. This process was applied by determination of best vision sphere and cylinder (magnitude and axis) using the Jackson cross-cylinder technique. 14 Owing to sphere power adjustment, variations in cylinder power were partly compensated; however, the entire compensations were subjectively controlled as well. For further measures, the cylinder highest power-that requested to be enhanced----was determined. After monocular subjective refraction, binocular balance was applied by alternate occlusion. The alternate occlusion method was performed using a handheld cover paddle. Both eyes were fogged with +1.00 D at the end of monocular subjective refraction and individuals were told to look at an isolated line of Snellen chart (6/9). Consequently, individual was asked to compare vision clarity between the right and left eyes while each eye was occluded for about half a second. If the individual confirmed a clearer vision in one eye, +0.25 D was added to the eye with clearer vision and alternate occlusion was repeated until equality was approached. The endpoint was the equal clarity in both eyes. Then, the ultimate spherical power was defined as the highest plus value or the lowest minus value that provided the best visual acuity. Three consecutive predicted phoropter refractions (PPR), were performed using the Zywave aberrometer (Bausch & Lomb, Rochester, NY). PPR is estimated using a second-order only Zernike model to fit the higher-order wavefront data for a nominal pupil diameter of 3.5 mm or in a specific exam pupil diameter. The Zywave aberrometer is a Hartmann-Shack wavefront sensor using wavelength of 785 nm. The accuracy of this device in measuring aberrations have been reported in a previous study. 15 The measurements were immediately perfomed after a blink to minimize disruption of the tear film. Fogging of the accommodative target was turned off during measurements with this device. A 6-mm optical zone was chosen for wavefront analysis. Dry (noncycloplegic) autorefraction was done using an autorefractometer (KR-800; Topcon, Tokyo, Japan) by automatically averaging the three measurements taken at the central 3-mm of the entrance pupil. Cycloplegic effect was obtained using cyclopentolate hydrochloride 1%. Three drops of this agent were administered with a time interval of five minutes. Thirty minutes after the third drop, autorefraction was repeated under cycloplegic conditions. Refractive data were recorded as sphere, negative cylinder, and cylinder axis. The astigmatic difference between the four methods was determined using the vector components of refraction. For this purpose, the conventional refraction was transformed to Jackson crossed cylinder format to calculate M, J 0 , and J 45 from the refractive data (S: sphere, C: cylinder, and ␣: cylinder axis) as the following: To compute the astigmatic difference between subjective refraction and the other three methods, the difference in J 0 and J 45 was calculated separately. Finally, the final values obtained of power vector components were transformed into the conventional cylinder using the following formula for a better understanding of the outcomes: All measurements were carried out between 4 pm and 8 pm. Cycloplegic autorefraction was always performed last. Selection of the other methods for each subject (subjective refraction, wavefront-based refraction and dry autorefraction) was done randomly. Refractive data was collected from both the right and left eyes of each subject, however, it is extremely likely that separate eyes of an individual would not reflect distinct samples and that there would be a significant association between refractive error components of the right and left eye, 16 and between higher orders aberrations of the right and left eye. 17 Thus, the right eye refractive data were recorded for the analysis. Data were analyzed using the SPSS.21 software (SPSS Inc, Chicago, Illinois, USA). Normality of the data was assessed using the Kolmogorov---Smirnov test, which did not show a normal distribution for quantitative data. The non-parametric equivalent of repeated measures analysis of variance (Friedman test) was used to compare the mean sphere, cylinder and spherical equivalent (SE) obtained with the different methods and the Bonferroni adjustment applied for pairwise comparisons. Correlation of objective refraction measures with subjective refraction as the reference standard was assessed using the Pearson correlation test. Bland-Altman plots 18 and intraclass correlation coefficients (ICC) were used to assess the agreement of subjective refraction with the other methods. ICC values of 0−0.2, 0.3−0.4, 0.5−0.6, 0.7−0.8 and more than 0.8 were considered indicators of poor, fair, moderate, strong and near perfect agreements, respectively. 19 The limits of agreement (LoA) were determined as the mean difference±1.96SD of the mean differences. In all tests, pvalues less than 0.05 were considered significant. Results Of the 83 subjects studied, 27 (32.5%) were males and 56 (67.5%) females. The mean age was 28.27 ± 6.87 years, ranging from 19 to 57 years. Table 1 summarizes the sphere, cylinder and spherical equivalent outcomes obtained using the different methods evaluated. The highest negative spherical, cylindrical and SE components were obtained using dry autorefraction, wavefront-based refraction and dry autorefraction, respec-tively. The lowest negative spherical, cylindrical and SE components were obtained using cycloplegic autorefraction, subjective refraction and cycloplegic autorefraction, respectively. Comparison of the repeated refractions using

different techniques showed statistically significant difference for all components using Friedman's test. The Dunn-Bonferroni post hoc test was used for the pairwise comparisons. The results are given in Table 1. The mean difference and correlation of the refractive components obtained using different techniques in comparison with subjective refraction as the standard reference is shown in Table 2. Considering the spherical component, there was a statistically significant hyperopic shift (0.12 ± 0.29 D, p = 0.001) with cycloplegic autorefraction compared to subjective refraction, and a significant myopic shift (−0.17 ± 0.32 D, p < 0.001) with dry autorefraction, while the difference between wavefront-based and subjective refraction was not significant statistically (p = 0.145). The calculated cylindrical component using subjective refraction showed statistically significant differences with dry autorefraction (p < 0.001), cycloplegic autorefraction (p = 0.041) and wavefront-based refraction (p < 0.001). As Table 2 shows, the lowest mean difference with subjective refraction in the obtained spherical and cylindrical components were found for wavefront-based refraction and cycloplegic autorefraction; however, the highest agreement was seen between cycloplegic and dry autorefraction, with narrower agreement bands (±0.58 and ±0.25 D) compared to the other methods. Considering the spherical equivalent, the lowest mean difference of SE (0.03 ± 0.39) and the highest agreement (±0.52) with subjective refraction were related to wavefront-based refraction and cycloplegic autorefraction, respectively. This good agreement was confirmed after obtaining higher values of ICC of cycloplegic refraction. Also, there was statistically significant correlation for all three refractive components (sphere, cylinder and SE) between all techniques with subjective refraction. The highest correlation with subjective refraction in sphere, cylinder and SE was observed for cycloplegic autorefraction (r s = 0.967), dry autorefraction (r s = 0.983) and cycloplegic autorefraction (r s = 0.982), respectively. According to the high and statistically significant correlation between the refractive components obtained using various techniques with subjective refraction, the Bland-Altman (1986) plot is a suitable method for analysis and comparison in situations in which different techniques measure an identical parameter. In these plots, the mean difference for SE between the two techniques is shown as a function of the mean SE of the two techniques. In each plot, the mean difference between each method with subjective refraction was marked with a horizontal solid line and the 95% limits of agreement with horizontal dashed lines. (Fig. 1, 2 and 3---3) In all plots, more than 95% of the differences between the SE determined with each technique and subjective refraction were placed in the range ±2SD of the mean difference in SE with the two methods. The lowest agreement band or the highest agreement with subjective refraction was seen for cycloplegic autorefraction. Also, the highest percentage of zero mean difference (15.7%) was found when subjective refraction and cycloplegic autorefraction were compared. Discussion The present study shows that sphere in cycloplegic autorefraction and astigmatism in dry autorefraction is in better agreement and correlation with the results of subjective refraction. The highest agreement with subjective refraction in SE was seen for cycloplegic autorefraction. Nearly 81.9% of the obtained SE with cycloplegic autorefraction, 67.5% of the results using dry autorefraction and53.1% of the results with wavefront-based refraction were within ±0.25 diopters of subjective refraction values. Compared with sphere of subjective refraction, there was a statistically significant hyperopic shift of about 0.12 D of sphere with cycloplegic autorefraction and a significant myopic shift of about 0.17 D with the value obtained with dry autorefraction in most of patients, while the difference between wavefront-based refraction and subjective refraction was not statistically significant. According to cycloplegic refraction, some patients showed a myopic shift. There are several explanations for this phenomenon: Incomplete cycloplegia and the increased accommodative effort may be considered as a causative factor; however, we tried to minimize this error by providing sufficient time for cycloplegia, drug administration by the examiner and determining refractive error by the same optometrist with residual accommodation ≤0.5 D and the same instrument. Moreover, spherical aberration may be responsible for the myopic shift of refractive error. Some studies reported that spherical aberration increased significantly following cycloplegia. 20,21 Gao et al. mentioned a significant reduction in the crystalline lens thickness following cycloplegia and backward lens shift which increased spherical aberration. 22 Therefore, with a mydriatic pupil, the crystalline lens' periphery is engaged in retinal image creation, which rises a positive spherical aberration and moves the image focus towards an anterior plane. 23 In agreement with the current study, it was reported that dry or non-cycloplegic autorefraction may lead to overminus or under-plus refractive outputs. 24 Giessler et al. showed aberrometers tend to underestimate myopia, 25 which is contrary to the previously reported instrumental myopia induced by most of aberrometers. 4,26 Thibos et al. suggested that aberrometry may eventually become the new standard for optimal correction of refractive error, 27 although aberrometry can provide a tremendous amount and detailed data on total aberration and each component of the Zernike polynomial decomposition of the wavefront error. For the comparison of objective refraction methods, some influencing factors have to be considered. First, the spherical and cylindrical lenses chosen by the patient may compensate for not only the spherocylindrical (lower order) aberrations, but also part of their higher orders aberrations. 28,29 It is known that coma can be compensated for, in part, by a cylindrical lens and spherical aberration by a spherical lens in a particular pupil size. Higher order aberrations may influence manifest refraction, and can be compensated for by using different spherical and cylindrical combinations when measuring subjective refraction, 27,30 while wavefront measurements split clearly between true spherical/cylindrical components and other higher orders aberrations. According to the types of higher order aberrations, spherical or cylindrical components may differ from the subjective refraction. Second, the pupil sizes during measurements is another possible factor explaining our outcomes. 31 A 3-mm smaller region of the pupil was used during the autorefraction while a 6-mm optical zone was usually chosen when using the wavefront-based refraction. Third, accommodation during measurement (instrumental myopia) could play a role when not dilated. There are other studies which found more myopic refraction using different aberrometers, and the instrumental myopia was proposed as one possible explanation for such findings. 5,15,32 Fourth, in evaluating the refractive data, the aberrometers' repeatability should be considered. In investigations on human as well as model eyes, some studies have shown acceptable repeatability of aberrometric measures, 12,33---37 while others demonstrate short-term variations in aberrations secondary to variations in accommodation, 38---40 small fixational eye movements 40 or even fluctuations in the tear film. 39,40 Fifth, wavelength is another parameter which can contribute to a potential artifact in the measurement of aberrometry. 35,41 The infrared (IR) light used in wavefront 42 Martin et al. 43 revealed in a study using visible monochromatic light (633 nm) that objective refraction compared to the subjective results were more hyperopic. Sixth, age was repsorted as a possible influencing factor according to the age-related reduction in the the foveal pit depth, 44 which results in less impact from the anterior vitreoretinal interface. All these fluctuations together with any illness process, can change the relative place reflecting the light used to select the target refraction in the aberrometric measurement procedure. The present research only included one subject of age above 50 years; therefore, the influence of this factor can be assumed to be very limited. Comparing with the subjective refraction as a reference, the astigmatism measured by dry autorefraction was more precise than the values obtained with the other methods, although wavefront-based refraction and cycloplegic autorefraction showed good correlation with subjective refraction. Astigmatism components have been generally found to be very similar among wavefront-based and subjective refractions. 27,32,45---48 In astigmatism, the correlation between wavefront-based and subjective refraction was not as high as the correlation obtained among the other with the two methods. Pesudovs et al. also reported that wavefront-based refraction were not as precise as standard autorefractions in estimation of astigmatism, although not clinically significantly worse. 49 Sphere from cycloplegic autorefraction and astigmatism from dry autorefraction are useful objective clinical data that can used as starting points for subjective refraction for most patients. A point that should be clarified is that refraction and prescription imply distinctive concepts. Prescription refers to a thinking process involving various aspects, including previous prescription, the probability of a new prescription being endured and patient needs. 50,51 Therefore, it is concluded that refraction alone fails to be a good alternative for prescription, irrespective of which type of refractions are concerned. One limitation in the current study was the assessment of the refractive status by an examiner, which may be a source of bias in measurement. It is recommended to repeat the same study in subjects with a previous history of corneal refractive surgery or in cases with irregular corneas. Conclusion Cycloplegic autorefraction in sphere and noncycloplegic autorefraction in astigmatism give a better estimate of subjective manifest refraction. Considering the subjective refraction as the gold standard in correcting refractive errors or modifying the present correction, these findings are very important in situations where subjective refraction is impossible, for example in infants, uncooperative and uncommunicative patients. Both autorefraction and aberrometry results, in general, showed a high level of agreement with subjective refraction. Relationship Between Nutritional Status and Students’ Physical Condition The issue in this research is the low level of FIK UNP students’ physical condition. The aim of this research to determine the relationship between nutritional status and students’ physical condition. This is a correlational research design involving 17 students of FIK UNP. Nutritional status data were obtained by measuring the Body Mass Index (BMI), and students’ physical condition data were obtained through test in several aspects including speed, strength, explosive power and endurance. Data were analyzed using Pearson’s product-moment correlation. The results showed that the average of students’ nutritional status was at 19.81 with normal classification, while the average of students’ physical condition was 56.01 with good classification. Based on the statistical analysis, it was obtained that tobserve 1.79> ttable 1.753. Therefore, it can be concluded that there is a significant relationship between nutritional status and FIK UNP students’ physical condition. INTRODUCTION Faculty of Sport Science Faculty organizes an education in the field of sports in which the course is largely practical subjects. In undergoing the activity as a student at the Faculty of Sport Sciences students are required to have excellent physical condition because the students will be busy with a variety of activities ranging from lectures to activities outside the lecture course, consuming energy. However, from the observations made by the author, is often seen student experience fatigue when the lecture started, and the number of students who are not eager to take the classes, especially the course of practice, this will certainly be inhibit student of activities that should be followed every day. The number of students still experience significant fatigue at the time of a new lecture and low morale likely caused by poor physical condition which is owned by the student because the physical condition of a component that can affect all human activities. Peak physical condition can make a person can carry out all activities to the maximum without experiencing significant fatigue, and can improve the spirit within themselves. Aside from the physical condition factor, the other thing that is likely to affect the students in activities and affect the physical condition itself is a factor of the nutritional status of every student. Nutritional status is a nutritional state that reflects the nutritional intake of a person. The nutritional status can affect the state of the human body, if the nutritional status is good, it will cause physical condition will be better because of insufficient supply of energy needed by the body to support all activities of daily living. Therefore, researchers interested in conducting research on the relationship between nutritional status and the ability of the physical condition of the results of this study are expected to provide an overview of the state of nutritional status and physical condition of the students and can provide solutions to the problems faced by some students of the Faculty of Sport Sciences, State University of Padang Nutritional Status "Nutritional status is the level of a person's health is influenced by food eaten that their physical impact can be measured anthropometry" [1]. Furthermore Syafrizar (2009: 4) says nutritional status is "state of

the body as a result consumption, absorption and utilization of food in the body". From the above opinion, it can be said that nutritional status is a state body that is affected by the consumption of food. The nutritional status is important for everyone's business because of the advantages and disadvantages of each nutrient can cause many disturbances of health. There are some advantages and disadvantages of the effects of nutrients, namely: "(1) a shortage of carbohydrates cause blood glucose is low, resulting in sluggish and lazy; (2) The resulting protein deficiency and meramus kwashiorkor. Excess protein lead to more severe kidney function and liver disease, as well as a decrease in bone calcium. (3) excessive fat reserves in the body resulting in some heart diseases, kidney, diabetes, high blood pressure, digestive disorders and other diseases [2]. Basedon the above quote can be interpreted that the carbohydrate, protein, calcium and fat is essential for health. Lack of carbohydrates will cause the body to become sluggish. Whereas if you eat protein deficiency will cause liver disease, kidney function will be heavy. While the calcium needed for bone growth and bone strengthening, if there is a deficiency of calcium will cause bone growth is inhibited and it will be easier brittle bones. Next need fat. Fat needed by the body as a source of energy, as a source of cell growth, support brain function, helps the absorption of vitamins and support the health of organs. Likewise, if our body fat reserves in excess, will cause obesity will lead to heart disease, kidney, diabetes, high blood pressure, digestive disorders and other degenerative diseases. In addition to carbohydrates, protein and fat or balanced, the body also needs vitamin, for example: Vitamin A serves to help the vision, help growth, maintain healthy skin, maintaining the respiratory system and protects the body from infection. Vitamin B 1 serves to maintain the health of the mucous membranes, maintain the nervous system, muscles and heart. Vitamin D functions for growth, amplification and maintenance of bones with calcium absorption to help the formation of strong bones and teeth, maintaining healthy blood, muscles and nerves. Deficiency or excess of vitamins can cause health problems body. The nutritional status is very dependent on the eating habits of everyday life, because nutrition is sufficient if not accompanied by a good diet, especially eating schedule will negatively impact the body, therefore it is important for students of the faculty of Sport Science to pay attention to nutrition nutrients from consumed food and eating well arranged so that nutrition can be met so as to support the daily activities as a student. Someone who experienced nutritional deficiencies. Physical Condition "The physical condition of the sport is all physical abilities which determine achievement whose realization is done through personal ability" [4]. "The physical condition is one of the preconditions that are necessary in any efforts to improve the achievement of an athlete, even said to be the foundation of the starting point of a prefix sporting achievement" [5]. Meanwhile, general physical condition can be interpreted by a state or physical ability [3]. From the opinions above, it can be understood that the physical condition is the ability of the body. The physical condition comprises: Speed "The speed is the ability to move with the fastest speed possibility" [8]. While is the speed of the body's ability to direct all of its systems in the fight against the load, distance and time produce mechanical work [3]. Of the two opinions, it can be said that the speed is the body's ability to travel long distances in a short time. Strength The strength of muscle contraction force is achieved in a maximum effort [8]. According in "power describes the ability of muscles to contract maximally yangh generated by a muscle or group of muscles" [9]. So power is the maximum capacity when the muscles contract. Explosive Power Explosive power is a blend or a combination of strength and speed, and "the explosive power is the ability of muscles to overcome the resistance of the load at high speed in a movement intact' [2]. So explosive power is the ability of muscles to be able to move strongly and quickly. Endurance Endurance is the ability to overcome fatigue and psychological perform physical labor for a long time [3]. According endurance mean length of a person to perform physical activities without experiencing significant fatigue [2]. So we can say that to endurance is the body's ability to cope with the fatigue when doing physical activity. The physical condition of a human needs to be able to meet all their daily activities, especially sport science faculty students who have a physical activity so much that it takes peak physical condition for all activities that followed can be accomplished smoothly. METHOD This type of research is a correlation study that looked at the relationship of two or more groups of concern in this study is the relationship between nutritional status (variable X) with the ability of the physical condition (Y). The samples in this study were students of the Sport Science Faculty, of Sport amounting to 17 people. Nutritional Status Based on measurements of nutritional status has been conducted on a sample by using Body Mass Index (BMI), then gained an average of 19.81, median 20:05, 20:20 mode, standard deviation 1:54, 22:23 highest score and lowest score 15.63. for more details of the nutritional status of the measurement results can be seen in the From table 1 above it can be seen that the first students (5.88%) had the nutritional status of score <17.0 with the classification of underweight weight level, 1 person students (5.88%) has nutritional status at the interval 17.0-18.5 by category underweight light level, and 15 students (88.24%) had a normal nutritional status categories and no student who has the nutritional status of overweight light level overweight and weight levels. Thus it can be said that most of the students of the Faculty of Sport Sciences sampled in this study had nutritional status on the score of 18.5-25 with normal categories and a percentage of 88.24%. the data is then presented into diagram 1. Physical Condition Based on measurements of the physical condition that has been done on samples obtained an average of 56.01, 58.39 median, standard deviation of 6:51, the highest score of 64.58 and the lowest score 43.20. for more details of the physical condition of the measurement results can be seen in table 2: Than 2 above tabel it can be seen that no student (0.00%) had the ability physical condition scores <65.78 with very good category, at 59.27-65.78 score as many as 6 people (35.29%) in good categories, 6 students (35.29%) had the physical conditions at the interval 52.76-59.26 with average category, and 3 students (17.65%) had a physical condition in the poor category and 2 students (11.76) which has a physical condition at very poor category. Thus it can be said that most of the students of the Faculty of Sport Sciences sampled in this study has the capability of physical conditions in both categories and are with a percentage of 35.29%. these results can also be seen in the diagram 2. Hypothesis Analysis Hypothesis in this research is a significant relationship between nutritional status and physical condition of the ability of students of the Faculty of Sport Sciences. Hypotheses were tested using correlation test and continued with t significance test at significance level α of 0.05%. From the analysis of the data obtained bytthe (1.79)> ttable (1,753), and thus H0 is rejected and Ha accepted meaning this hypothesis empirically accepted as true. Based on the research that has been done it can be seen that the hypothesis that there is a significant relationship between nutritional status and the ability of the physical condition can be accepted as true empirically. This proves that the relationship between physical condition and nutritional status directly proportional, if the nutritional status that are owned by either the student will come good physical condition, but on the contrary if the nutritional status of which is owned by the student is low, then the state will participate fisikpun low. Thus, one of the businesses that can be done to improve the physical condition of students is to improve the nutritional status of students themselves. For improving the nutritional status it must first know the factors that could affect the nutritional status. The factors affecting the nutritional status of a person, namely: "1) Food products, 2) distribution of food or food, 3) acceptability (acceptance), 4) the prejudice in certain foodstuffs, 5) prohibition on certain foods, 6) preference for certain foods, 7) economic constraints, 8) eating habits, 9) appetite, 10) sanitation of food (preparation, presentation, storage) and 11) knowledge of nutrition" [6]. Furthermore "Factors affecting the nutritional status, among others, (a) the availability of food; (b) knowledge of nutrition; (c) eating habits; (d) the level of income" [7]. Of the two opinions are the things to consider in improving the nutritional status is by paying attention to food availability. Food security is a guarantee of the availability of raw materials needs to be consumed in order to avoid gaps or lack of eating that has enough nutritional value. Next is to provide education or knowledge to students to pay attention to the nutritional value of food consumed every day and observe the variety of foods consumed, the next is adjust your diet, diet is crucial once the benefits of nutritional consumed, if wrong in making dietary adjustments the nutrients produced by the food consumed will not be helpful to the maximum can even cause harm such as the accumulation of fat, diseases of the stomach as well as a variety of other adverse effects. Therefore, in order that the nutrients produced by the food is consumed can provide the maximum benefit would require setting a good diet. CONCLUSION Based on the research that has been done it can be concluded in this study is a significant relationship between nutritional status and physical condition of the ability of Sport Science Faculty student Padang State University. Based on the conclusions set forth above, the authors give advice to students in order to pay attention to the nutritional value of food dikonsumsi and regulate eating well. Measuring medically unjustified hospitalizations in Switzerland Background Inappropriate use of acute hospital beds is a major topic in health politics. We present here a new approach to measure unnecessary hospitalizations in Medicine and Pediatrics. Methods The necessity of a hospital admission was determined using explicit criteria related to the recorded diagnoses. Two indicators (i.e. “unjustified” and “sometimes justified” stays) were applied to more than 800,000 hospital stays and a random sample of 200 of them was analyzed by two clinicians, using routine data available in medical statistics. The validation of the indicators focused on their precision, validity and adjustment, as well as their usefulness (i.e. interest and risk of abuse). Results Rates, adjusted for case mix (i.e. age of patient, admission planned or not), showed statistically significant differences among hospitals. Only 6.5% of false positives were observed for “unjustified stays” and 17% for “sometimes justified stays”. Respectively 7 and 12% of stays had an unknown status, due to a lack of sufficiently precise data. Considering true positives only, almost one third of medical and pediatric stays were classified as not strictly justified from a medical point of view in Switzerland. Among these stays, about one fifth could have probably been avoided without risk. To enable a larger ambulatory shift, recommendations were made to strengthen the ambulatory care, notably regarding post-emergency follow-up, cardiac and pulmonary functions’ monitoring, pain management, falls prevention, and specialized at-home services that should be offered. Conclusion We recommend using “unjustified stays” and “sometimes justified stays” indicators to monitor inappropriate hospitalizations. The latter could help the planning of reinforced ambulatory care measures to pursue the ambulatory shift. Nonetheless, we clearly advise against the use of these two indicators for hospitals financing purposes. Supplementary Information The online version contains supplementary material available at 10.1186/s12913-022-07569-3. Background Inappropriate use of acute hospital beds is a challenging issue that has several detrimental implications. Not only does it increase the health care costs (in Switzerland, a group of experts recently highlighted the importance of the ambulatory shift, to reduce

these costs [1]), but hospitalization in itself might be harmful [2,3]. In Switzerland, except for some surgical procedures only reimbursed as ambulatory care (varicose veins, hemorrhoids, menisci, etc.), hospitalizations are always reimbursed, regardless of whether the admission was justified [4]. Whereas several authors analyzed the potential move towards ambulatory surgery [5][6][7][8], little has been proposed to substitute hospitalizations by ambulatory care, especially in medicine and pediatrics. To reduce the number of hospitalizations one can either prevent them by offering optimal ambulatory care or avoid unjustified ones. The Ambulatory Care Sensitive Conditions (ACSC) indicator enables to monitor the first aspect by screening stays that might have been prevented had the patient had access to adequate primary care services [9,10]. This indicator has the advantage of being very easy to compute from routinely available hospital medical records. However, it suffers several limitations, including lack of sensitivity and specificity [11], poor clinical relevance [12], and low proportion of hospitalization actually preventable [13], its main drawback being that high rates of ACSC might be associated to differences in admission hospital practices rather than to the quality of ambulatory care [14]. To work on the second aspect of hospitalization's reduction, one can use the Appropriateness Evaluation Protocol (AEP) [15], which focuses on the justification of hospital admissions. Despite several adaptations, this protocol fails to provide consistent rates of inappropriate hospitalizations [16][17][18]. It is based on a limited number of conditions related to patients and does not account for newly available alternative ambulatory services, such as outpatient intravenous therapy, home oxygen therapy, 24 h nursing care services, and home rehabilitation. Moreover, when confronted to an expert panel consensus, AEP's accuracy has only been judged as "fair" [19]. These limitations and the fact that AEP requires a detailed and labor-intensive review of medical records might explain why the use of this protocol, which was extensive in many European countries during almost three decades [20], has decreased lately. Another approach, based on routinely available hospital data, consists in identifying health conditions -defined by ICD 10 -that could be managed without emergency admission to an inpatient bed [21] (about one fifth of all emergency admissions in England [22]). Unfortunately, this approach fails to account for secondary diagnoses or interventions, which might justify hospitalizations. Moreover, it ignores elective hospital admissions that can sometimes be substituted by ambulatory care (e.g., to investigate non-severe conditions) and, thus, prevented. In short, we are currently missing a tool to screen hospital admissions that are not justified from a medical point of view. Such a tool would enable to measure the frequency of ACSC without biases due to differences in admission policies. In addition, it would be useful to evaluate the share of hospitalizations that might be prevented by offering appropriate ambulatory care. An innovative approach has been proposed by SQLape ® classification of patients, which is based on multiple diagnostic and surgical categories [23,24]. If a patient has at least one diagnosis or one surgical operation justifying the hospitalization, their admission is considered justified. Excluding surgical and obstetrical stays (which are always considered as justified), all the other stays are considered poorly justified and classified either as "unjustified" (i.e. when all diagnoses "almost never" justify a hospital admission per se) or "sometimes justified" (i.e. when at least one diagnosis "sometimes" justifies a hospitalization depending on the severity of the illness) [25]. The categorization of diagnoses was established on an empirical basis over several years, with an adjustment of the algorithms through a feedback loop by a dozen of hospitals in the French and Italian parts of Switzerland. However, the indicators have not been scientifically validated yet, hence this paper. Using routinely available data, we assessed the strengths and limitations of the "unjustified stays" and "sometimes justified stays" indicators. Classically, a good indicator should be unbiased, precise and valid, i.e. adjusted for risk factors having a strong association with the outcome, providing statistically significant deviations among hospitals, without too much false positives or negatives (numerator), and with a proper eligible population (denominator) [26]. Baker and Chassin recently proposed to add two additional criteria to judge the usefulness of outcome indicators [27]: providers should be able to influence substantially the outcome and its use should have little chance of inducing unintended adverse consequences. The objective of our article was to present our two innovative indicators and to validate them according to the above criteria. We did not put too much emphasize on avoiding false negatives since the intention was to provide a measure of unjustified stays that could potentially be avoided without too much dispute (minimum value). This criterion could be strengthened if needed. We used Swiss hospital medical statistics to provide a validation based on representative and extensive results by hospital, as well a sample of hospitalization to assess the frequency of false positives from a clinical point of view. Data The source population included all hospitalizations recorded in the Medical Statistics of Hospitals (Federal Statistical Office) with a discharge occurring in the years 2014 to 2016 and a length of stay greater than 1 (i.e. different admission and discharge dates). Hospitalizations with chemo-or radiotherapy, with a surgical intervention requiring a surgical theater, or related to obstetrics (i.e. delivery, abortion) were excluded using the standard SQLape ® tool [23, 24]. Moreover, based on suggestions from the clinicians reviewing our tool, we additionally excluded: -stays with an admission after 6 PM and a discharge the next day, considering the time required to avoid missing a high risk; -newborn stays (less than one year), since it was impossible to evaluate the prognosis from the minimal data set; -and elective hospitalization for alcohol use disorder (ICD-10 Z502 or ICD-9-CM 9462 codes). Screening "unjustified stays" and "sometimes justified stays" A stay was considered "justified" if one of the following conditions was met [23]: -at least one diagnosis classified as "almost always" requiring a hospitalization (premature birth, acute myocardial infarction, shock, pulmonary embolism, stroke, peritonitis, agranulocytosis for instance); -more than two failures among following vital organs: respiratory (chronic respiratory failure), cerebral (degenerative disease of brain, dementia), cardiac (heart failure), hepatic (liver cirrhosis), hematologic (coagulation disorders), renal (chronic nephropathy, end stage renal disease) [28]; -cardiac dysrhythmia with cardiac failure; -hallucination or delirium of a patient not living in a nursing home; -pneumonia if children less than 7 years old or patients with significant comorbidities (cardiac congenital malformation, heart failure, other disease of large vessels, interstitial pulmonary disease, acquired immunodeficiency syndrome, other immune disorder). A stay was considered as "unjustified" if all its associated conditions were classified as "almost never" requiring a hospitalization (Parkinson's disease, migraine, anemia, psoriasis, thyroid disorders for instance). All other stays were classified as "sometimes justified" (called "more or less justified stays" in the SQLape ® tool [24]). The list of conditions that "almost always", "sometimes", or "almost never" require a hospitalization is provided in Additional file 1. Review of cases screened Two senior clinicians (PH, TB) reviewed independently a random sample of 200 "unjustified" hospitalizations and another one of 200 "sometimes justified" stays screened by the SQLape tool. PH has a long experience in hospital and ambulatory settings as internist. TB was in charge of academic and medical supervision of the unit of family medicine in the Primary Care and Public Health Center (Unisanté). Available data were age, gender, hospital department (medicine, surgery, pediatrics, obstetrics, etc.), length of stay, main and secondary diagnoses, procedures and delay between admission and procedure dates. Data were printed as a case summary and presented in one or two pages by hospital stay. All data were anonymous and did not include any information enabling the identification of the individuals (no date of birth, ZIP code, hospitals, etc.) [29]. In a first round, the two senior clinicians reviewed all the stays independently and had to answer the two following questions: • Question 1: From your point of view, was this hospital stay necessary in an ideal context? Possible answers: yes, no, I don't know. • Question 2: If the stay is not necessary, give the most probable cause of the hospital admission. For question 2, the reviewer had to choose among an a-priori list of reasons that might have precluded home return (listed in Additional file 2). In a second round, all divergent judgments on the necessity of admission were discussed in depth until a consensus was reached or a decision to keep the divergence was taken. The aim of this discussion was to make sure both clinicians considered all possible situations corresponding to the combination of diagnoses and treatments. They were authorized to reach a consensus "I don't know" if they both judged that the divergence was due to a lack of more detailed data (e.g. labs results, medication or social environment). Statistical analysis Following de Mast [30] and Brennan [31], we assessed agreement between the two reviewers' judgment (three categories: necessary, unnecessary, unknown) using uniform kappa rather than Cohen's kappa [32]. The uniform kappa assumes a chance measurement independent of the categories being measured and, thus, a probability of agreement by chance constant and equal to the inverse of the number of categories (here 0.33). Since the outcome from second round was obtained by consensus (i.e. no independency), we did not report kappa and simply computed the agreement percentage (i.e. number of identical judgements divided by the total of observations). Observed, expected rates and surplus estimation Observed rates were obtained by dividing the numbers of cases (i.e. unjustified or sometimes justified stays) by the number of eligible stays as defined above. Using data from years 2014-2016, we computed expected rates by strata of risk, i.e. by age classes and type of admission (programmed vs urgent), in a two steps procedure. First, we computed rates on the whole sample, which contained all Swiss hospitals. Second, we selected hospitals with observed rates lower than expected ones, among general health care hospitals (referred as K1-type hospitals [33]). Third, we computed new average expected rates based on these benchmark hospitals. We applied control limits to take into account the random variations of both observed and expected rates [34]. Outlying hospitals correspond to hospitals with observed rates exceeding maximal control limits at the 95% level (unilateral test). Since hospitals may require more than 24 h to exclude a life-threatening condition or serve social needs (e.g. child abused protection), the target of avoiding all "unjustified" and "sometimes justified" stays is not desirable. Consequently, we estimated the surplus as the difference between observed and expected rates multiplied by the size of the eligible population. Results In the first round, higher kappa was obtained for unjustified than for sometimes justified stays (0.70 vs 0.50). After discussion (i.e. after round 2), 86% of stays screened as unjustified were classified as unnecessary by both reviewers (Table 1). This proportion was slightly lower (71%) for "sometimes justified" stays. About 7.5% of "unjustified" stays remained difficult to evaluate (i.e. unknown status for both reviewers) and about 5.5% were false positive (i.e. necessary for both reviewers). For "sometimes justified" stays, these proportions were a bit higher (11.5% unknown; 16% false positive). The reasons for precluding home return are given in Table 2. A quarter of unjustified stays had no severe diagnosis after investigation, but patients were admitted as a precaution. About 20% of patients were admitted for pain treatment. The third most frequent cause was lack of security at home (13.5%). Unjustified and sometimes justified stays represented 17.8 and 24% of eligible admissions, respectively (Table 3). Proportions were higher among children and elective hospitalizations. Globally, the unjustified surplus represented 2.5% of the eligible stays against 3.3% for the sometimes justified one, leading to a total of about 6% of poorly justified stays. We plotted the ratios of observed divided by expected rates for unjustified and sometimes justified stays against the number of eligible discharges (Figs. 1 and 2). Only hospitals with at least one admission per day are shown. Control limits are given by grey lines, hospitals with observed rates significantly higher than maximum expected ratios are represented by (red) triangles, hospitals with lower values in (green) circles. Since the expected rate was computed using benchmark hospitals only, a majority of hospitals have higher rates than expected (i.e. ratio > 1.0). There are as many hospitals above and below control limits among large and smaller hospitals for both indicators. Additional file 2

provides the list of all diagnoses involved in poorly justified hospitalizations, which might suggest some propositions for alternative ambulatory care (see Discussion below). Discussion In this paper, we applied the indicators "unjustified stays" and "sometimes justified stays" to a random sample of 200 hospitalizations in Medicine and Pediatrics to estimate the proportion of those hospitalizations that were indeed unnecessary from an a posteriori medical point of view and to understand the motives of those admissions. Then, we applied these two indicators to all eligible hospitalizations in Switzerland from 2014 to 2016 (820,000 stays) to analyze the variability of the results among hospitals and to estimate the potential of hospitalization reduction. Agreement between reviewers at the first round was rather good for unjustified stays, whereas it was only fair for "sometimes justified" stays. The main reason for these divergences was the lack of detailed clinical data, which lead the two reviewers to refer to different situations. The discussion of all possible stories improved the agreement Two clinicians made medical judgements and involving more physicians would probably result in less agreement. The proportion of unknown status increased between the first and the second round to about 8% for "unjustified" and 12% for "sometimes justified" stays (Table 1). These results emphasize the main limitation of our study: the lack of details on patients (e.g. no information about the severity of illnesses, laboratory, and drugs). It must be emphasized that the reviewers had much more precise information (more than 16,000 diagnostic codes and 11,000 operating codes) than the crude criteria used by the screening tool, which is based only on 200 diagnostic and intervention groups, respectively. The divergences did not concern grouping issues but severity, investigation, or treatment strategy aspects. Finally, having the reviewers formulating an opinion in about 90% of the cases is a strong argument that even if it would be preferable to have full access to the medical records in the hospitals, partial access should not invalidate the results. Overall, both reviewers estimated that the majority of screened stays were unnecessary: at least 86% for "unjustified stays" and at least 71% for "sometimes justified" ones ( Table 1). The false positive rate for "unjustified stays" (14%) is relatively low, allowing to use this indicator to push hospitals to be rigorous in their admission criteria. The false positive rate for "sometimes justified" remain acceptable, although too high to judge the performance of hospitals. It should thus be used more parsimoniously, for instance as a basis to the reflection on how to improve the health system. As mentioned in the introduction, we did not estimate the proportion of false negative, considering that this issue could be analyzed in further research. Our results were solely adjusted for patients' age and types of admissions (programmed or not), adjusting for this possible source of bias. Unmeasured factors like education or social characteristics might perhaps influence the rate of unjustified stays by hospitals but whether they should be introduced into the analysis is not so clear, since improvements (e.g. specific education services or social support to outpatients) might target such determinants. Both indicators provided precise estimations, with many hospitals having observed rates significantly above upper statistical control limits (Figs. 1 and 2). To answer the question of the usefulness of our indicators, we first analyzed the reasons of admitting patients, even if theirs stays were a posteriori unjustified for medical reasons. A quarter of unjustified hospitalizations was due to a conservative approach adopted by clinicians to rule out a high-risk diagnosis ("suspicion of serious illness", Table 2). A lot of these patients had diagnoses of non-specific disorders, pain, or psychiatric troubles (Additional file 2). Such stays might be avoided if accelerated diagnostic pathways were applied, as recommended for instance for chest pain [35], knowing that less than 10% of emergency department patients with chest pain are ultimately diagnosed with an acute coronary syndrome [36]. Such protocols were applied with evidence of being efficient for low-risk patients (early discharge) and high risk ones (early intervention or treatment) [37,38]. Another frequent reason for this kind of unjustified hospitalizations is related to the lack of outpatients' facilities to monitor serious illnesses. For instance, seizures or suspected seizures, which account for a large number of emergency admissions, might be prevented as suggested by geographical variability of admission rates [39]. Emergency care pathways might be applied to focus on rapid appointments in specialized services [40,41]. Other nonspecific complaints that could yield an unjustified stay for monitoring suspicious or serious illness include giddiness, cerebral disorders, and hypotension. Such conditions could benefit from clinical pathways' approaches, fast access to a specialist or brain imaging to rule out a brainstem lesion. Acute respiratory infections accounted for almost 6% of unjustified stays. We observed large variations among hospitals of admission rates for this condition, especially among young children, suggesting varying admission criteria. There is a substantial variation in the management of bronchiolitis and criteria of hospitalization or discharge to home are often subjective. Moreover, many admitted infants had no distress [42,43], raising the question of the continuity of care between ambulatory and hospital pediatricians for instance. The remaining causes of unnecessary stays are scarce, mostly due to contextual variables. For instance, isolation of immune-suppressed patient does not necessarily require hospital beds but might be difficult to obtain at home. Patients requiring investigation not available in ambulatory setting or living far from such infrastructure might prefer to stay in a hospital, though a hotel stay would perhaps provide the same comfort. Lack of compliance makes it difficult to find an alternative to hospitalization for instance for alcoholic, addicted people, or persons with intellectual disabilities. Care facilities at a lower level than general hospital beds, such as in nursing home, might also be offered in the proximity of patients' home, with a supervision by their primary care physicians. The analysis of the 820′000 hospitalizations in Swiss hospital of the period 2014-2016 showed variations of rates among hospitals. In the short term, public health services might ask hospitals not to exceed expected rates. In Switzerland, for the period 2014-2016, this would have yielded a reduction of about 6% of the hospitalizations (proportion of surplus of Table 3). To obtain a subsequent reduction of unjustified stays, several measures could be implemented, including: -encouraging hospitals to work more closely with outpatient facilities to identify faster patients with at-risk diagnoses and provide a secured monitoring (e.g. acute coronary syndrome, epilepsy); -providing community reinforcement of monitoring at home, implying home physician's and nurse's visits, education of patients and relatives; -supporting gradual and effective treatments for pain at home; -providing immediate home safety assessment and intervention rehabilitation to prevent dangerous situations (risk of falling, frail old patients); -pursuing the efforts to maintain patients at home, with more specialized home nurses' skills (IV antibiotherapy, parenteral nutrition, wound dressing or care; such services need a multidisciplinary approach to be successful [44]). The potential of reduction of the number of hospitalizations is substantial. Considering that about 18% of them were screened "unjustified", from which 86% were considered as unnecessary, and about 24% were screened "sometimes justified", from which 70% were deemed unnecessary, this yields a theoretical reduction target of approximately one third. Therefore, the short-term reduction (i.e. 6%) represents only 20% of the total potential reduction. In practice, this proportion should be considered as overuse only if less intensive care can provide similar outcomes. The question therefore arises whether it would be possible to reduce the number of hospitalizations without endangering patient safety and to what extent alternative inpatient care strategies should be tailored. Summarizing these results, we can conclude that hospitals might be able to influence the outcome and achieve a 6% hospitalizations reduction by themselves. The analysis however also provided some evidence that involving ambulatory care facilities would be necessary to achieve a more substantial ambulatory shift (up to 24% additional reduction of the number of hospitalizations). When using those indicators, one should be aware of the possible unintended adverse consequences. Justifying hospital stays based only on medical criteria can indeed lead to a possible harmful effect. Social or compassion care might be indicated if the hospital is the only place to shelter or surround a patient. Then, the medical justification of a stay is made a posteriori, without information about possible diagnoses considered at admission that might have justified a hospitalization. In addition, the accuracy of the indicator depends on the coding quality. For instance, if a severe acute respiratory insufficiency occurring during an influenza episode was not coded, the corresponding stay would wrongly be considered unnecessary. We therefore recommend analyzing the results carefully to see if a suboptimal coding quality might explain high rates. Finally, we discourage using these indicators to refuse funding of unjustified stays, since this could affect the security of care. Financial penalties might perhaps be used to encourage hospitals reaching the expected rates, but only globally (not for specific stays). Nevertheless, it must be kept in mind that ambulatory care also generates costs and that difficulties to improve the appropriateness of hospitalizations might also be related to regional aspects, such as insufficient ambulatory coverage. Although the frontier between ambulatory and hospital care is not universal, we believe that this study might be replicated and applied in other countries. Other classifications' tools could be used, given that co-morbidities are explicitly reflected and that diagnoses and intervention categories are sufficiently homogeneous to determine whether they justify hospitalizations. Some authors will probably propose refinements or adaptations (for instance, we had some difficulties to decide whether elective alcoholic withdrawals or non-traumatic painful back might justify hospitalizations). Further research should focus on the pediatric context to better understand interregional practice differences. Analyzing unjustified stays from detailed medical records is necessary to understand what kind of ambulatory care is missing to ensure secure alternatives to hospitalization. However, we believe that our results are interesting since they show that there is a substantial potential to shift toward ambulatory care in Medicine and Pediatrics and enable to delineate the most promising domains. In this respect, both indicators might be used for planning purpose. Conclusion We recommend using the "unjustified stays" and "sometimes justified stays" indicators to monitor inappropriate hospitalizations. Based on these two indicators, we found that one third of the medicine and pediatric hospitalizations made in Switzerland between 2014 and 2016 did not have a clear a-posteriori medical justification. Nevertheless, our results suggest that only a part of these stays (6% of eligible stays) could be avoided without changes in the health care system. To obtain a more substantial reduction whilst ensuring patients' safety, measures to reinforce ambulatory care are required. A Strategy Toward Reconstructing the Healthcare System of a Unified Korea This road map aims to establish a stable and integrated healthcare system for the Korean Peninsula by improving health conditions and building a foundation for healthcare in North Korea through a series of effective healthcare programs. With a basic time frame extending from the present in stages towards unification, the roadmap is composed of four successive phases. The first and second phases, each expected to last five years, respectively, focus on disease treatment and nutritional treatment. These phases would thereby safeguard the health of the most vulnerable populations in North Korea, while fulfilling the basic health needs of other groups by modernizing existing medical facilities. Based on the gains of the first two phases, the third phase, for ten years, would prepare for unification of the Koreas by promoting the health of all the North Korean people and improving basic infrastructural elements such as health workforce capacity and medical institutions. The fourth phase, assuming that unification will take place, provides fundamental principles and directions for establishing an integrated healthcare system across the Korean Peninsula. We are hoping to increase the consistency of the program and overcome several existing concerns of the current program with this roadmap. Current Status of Healthcare in North Korea The healthcare system in North Korea has degenerated, becoming an impotent system incapable of providing even minimal services to its citizens. Specifically, the integrated provision pISSN 1975-8375 eISSN 2233-4521 of services via a system of free medical care and "section doctors" appears to be non-existent, as most functions, including those performed by medical institutions, have been impaired [1][2][3]. The most pressing problem is that the government of North Korea

has failed to demonstrate willingness to address failures in its healthcare system. As part of a socialist system, the provision of resources depends solely on governmental effort, and it is impossible to innovate within the limits of the healthcare system of North Korea. Meanwhile, the North Korean people are in very poor health with a high disease burden, and the health outcome gap between South and North Korea is large and has been increasing over time [4]. The health and living conditions of children and women are particularly threatened, and non-communicable diseases have imposed an increasing burden on North Korea as the magnitude and severity of communicable diseases linger although the prevalence of communicable diseases has improved due to international contributions [5][6][7]. Ongoing Problems With Humanitarian Aid to North Korea To help North Koreans suffering due to the collapsed healthcare system, international organizations and the South Korean government have provided ongoing support to North Korea since 1995 [8,9]. However, the programs that have been implemented have faced several problems, such as an inadequate long-term budget, a low level of coordination among providers, and an inability to monitor the distribution system. Lack of a monitoring mechanism because of resistance by the North Korean government has been an important concern, as the international community has been unsure whether the support was being distributed to those in need. Another key issue is the role of participating agencies. The South Korean government underwrites most of the budget for many of the international health programs running in North Korea; however, South Korea has had a very limited role so far and has not developed any long-term plan or roadmap. The nongovernmental organizations (NGOs) that have initiated the current inter-Korean partnerships and have played a significant role in increasing trust, even in times of political and diplomatic tension, have only limited access to some areas in North Korea such as Pyeongyang and its suburbs. Necessity of a Long-term Roadmap for Healthcare in a Unified Korea From the perspective of South Korea, preparing for and seeking the unification of Korean peninsula, it is crucial to improve the health status of the North Korean people, which would contribute to the unification processes with maximized human security and to integrate two healthcare systems into a harmonized one, which is needed for more stable unification. At this point, we need new strategies to develop an effective support program for North Korea that can be incorporated into a long-term roadmap for healthcare system integration of the two Koreas while ensuring a systematic partnership among the participating agencies and stakeholders. In this article, based on the clear understanding of the situations mentioned above, we developed the following concise road map for healthcare programs in North Korea and to prepare for an integrated healthcare system between South and North Korea at the time of unification. Timeframe and Goals This road map aims to establish a stable and integrated healthcare system for the Korean Peninsula by improving health conditions and building a foundation for healthcare in North Korea through a series of effective healthcare programs ( Table 1). This roadmap is based on the scenario by the Ministry of Unification on "Building an infrastructure program for North Korea" which assumes unification in 10, 20, or 30 years [10]. This roadmap specifically focuses on the scenario of unification in twenty years. With a basic time frame ranging from the present in stages towards unification, the roadmap is composed of four successive phases. The first and second phases, each expected to last five years, focus on disease treatment and nutritional treatment, respectively. These phases would thereby safeguard the health of the most vulnerable populations in North Korea, while fulfilling the basic health needs of other groups by modernizing existing medical facilities. Based on the gains of the first two phases, the third phase, which would last for ten years, would prepare for unification of the Koreas by promoting health of all the North Korean people and improving basic infrastructural elements such as health workforce capacity and medical institutions. The fourth phase, assuming that unification will take place, provides fundamental principles and directions for establishing an integrated healthcare system across the Korean Peninsula. Key Strategy 1: Modernization of City/County Hospitals as the Base of Healthcare Services According to our scenario, due to the strained inter-Korean relationship, a limited range of programs would initially be carried out based on principles of selection and concentration, and gradually be expanded to comprise a nationwide omnidirectional program. In terms of content, the city and/or county would be established as a base for healthcare services, and the functions of city/county-level hospitals would be revitalized to implement and provide disease treatment, nutrition treatment, and health promotion services in North Korea. All cities and counties of North Korea have at least one hos-pital designated to serve the local people, and therefore, modernization of these hospitals can act as the key strategy for providing health services. Then over time, the number of modernized hospitals can be increased, and they will provide more services for more people. Key Strategy 2: Founding a New Institution as a Control Tower An institution provisionally named the 'North Korea Healthcare Foundation' must be founded to oversee the roadmap. Creating this institution would have several advantages: to maintain consistency of policy throughout the period; to maintain a steady relationship with international organizations working with North Korea; and to simplify and consolidate the collection of information about current North Korean healthcare issues. From phase 3, the institution should transition to having North Korean officials and experts as its members or staff because more a cooperative and nationwide program would be launched, and it will be better to change its name to the 'Korea Healthcare Foundation. ' Phase 1: Set up the Base Hospitals and Treat Those in Urgent Need (5 Years) Five city/county hospitals in each province would be selected, considering access to care. The selected 50 hospitals nationwide will provide life-saving care for those in urgent need including children, pregnant women, and those with contagious diseases. To serve the target population, selected hospitals will modernize with a focus on inpatient care and outpatient care facilities for pediatrics, obstetrics, and infectious diseases, and will perform prenatal care, delivery services, vaccination for children, and treatment of contagious diseases such as tuberculosis or malaria. Phase 2: Increase the Number of Base Hospitals and Add Other Specialty Services (5 Years) The number of base hospitals will be increased up to the number of all of the city/county hospitals, and this will improve the access to care dramatically. In addition, more specialized health services such as surgery and medication for chronic disease could be provided. For these services, hospital renovation will focus on operating rooms and sufficient supplies for operation. Patients with cardiovascular disease or diabetes will receive priority care because it is suspected that these chronic conditions are widespread in North Korea [6]. Signing an agreement between South and North Korea for healthcare services is a crucial goal by the end of this phase. Implementing a national level health enhancing program will be another important step. Phase 3: Run the Nationwide Health Promotion Program and Build the Healthcare Infrastructure for Integration (10 Years) Through improvement in the inter-Korean relationship and implementation of an inter-Korean agreement regarding healthcare, the third phase would provide an omnidirectional healthcare program to all citizens of North Korea as well as other basic healthcare/medical infrastructure. Phase 3 will focus on preventive care to minimize the health outcome gap between South and North Korea. Two main programs will be executed-a public education program for quitting smoking and binge drinking, and another to start exercise. A screening program to identify patients suffering from metabolic syndrome will be commenced when possible. For the preparation of healthcare system integration, in this phase, implementing more elaborate and systematic programs for enhancing the health workforce and medical facilities, which are the main resources of a sound healthcare system, will be needed. These tasks will be conducted efficiently through the renovation of the tertiary hospitals, including medical schools, which are responsible for medical education and general guidance to the lower level institutions. Phase 4: Integration for One System Complete integration of the healthcare system between the South and North is the goal. A national health insurance system in a specific area will be operated as a test for a smooth transition. In the beginning, the North Korean healthcare system will be preserved to some degree and South Korean pro-fessionals or hospitals might gradually start to operate in North Korea. South Korean Government It is important to create the North Korea Healthcare Foundation as early as possible. The North Korea Healthcare Foundation will control the whole process throughout the period and will partner with international organizations and NGOs. The roles of the North Korea Health Care Foundation are the selection of the base hospitals and the operation of training programs for North Korean healthcare professionals. During phase 1 and 2, the North Korea Healthcare Foundation will work closely with international organizations or NGOs that have experience working with North Korea and build a stable relationship with its North Korean counterparts. As phase 3 begins, the North Korea Healthcare Foundation will take the initiative for all of the programs by using public and private funding available in South Korea to run the programs. International Agencies and Korean Nongovernmental Organizations The World Health Organization (WHO) will run a hospital modernization program as an infant and children program [9], and the North Korea Healthcare Foundation will be a partner with the WHO to maximize their experience. Dividing roles among international organizations and NGOs is desirable. NGOs have limited access while international organizations have relatively unrestricted access to the whole country. If international organizations cover areas not accessible to NGOs, it will maximize the equal exposure of the program throughout North Korea. Under this scenario, NGOs will cover Pyeongyang and its suburbs, and the WHO will cover other areas in the northeast of the country. Currently the WHO and the United Nations Children's Fund run a program for those with tuberculosis and malaria throughout North Korea, funded by a global fund and a world vaccination association. Parasitic diseases will be treated by both preventive and curative methods. The South Korean government will provide parasite pills for all students and educate them to prevent possible infection. For the general population health bases, they will provide a major contribution for the control of parasitic diseases, and it is likely that NGOs will have access to Pyeongyang and its suburbs. The advantage of using Korean NGOs as opposed to international organizations is that there are relatively similar cultures and no language barrier. Hopefully, the NGOs will develop a good partnership with North Korea during the program and can then execute the program independently with North Korea later. The creation of a consortium of Korean NGOs would be advisable. CONCLUSION This roadmap includes step-by-step procedures needed for an integrated healthcare system between South and North Korea and detailed strategies for each phase to overcome several existing issues with humanitarian health programs in North Korea. With this long-term roadmap, we are hoping to increase the consistency of the program and setting up a clear role for each agency would minimize coordination issues. The South Korean government can perform the role of manager of the whole North Korea support program based on this roadmap. Out of several important considerations for the roadmap, the current and future political relationship between South and North Korea is the most crucial one. The roadmap can work properly only if North Korean authorities accept the plan and cooperate with sincerity, but due to the current tension in the inter-Korean relationship, it is difficult to even initiate such a long-term project. In the context of such an unpredictable relationship, a coherent and planned program aligned with clear principles like this roadmap is needed more than ever. For the acceptance and implementation of an effective healthcare program in North Korea, the government of South Korea must consistently persuade relevant stake holders, including the government of North Korea. At the starting point, it would be desirable for international agencies to take the initiative to launch and conduct this plan like the program they are currently running for women and children in North Korea. In this

process, such a 'planned' roadmap is a necessity, and our proposed roadmap would allow South and North Korea to both play important roles cooperatively in controlling the entire process of implementing an effective health program and improving the healthcare status of North Korea. Through this, the two Koreas would be one step closer to unification. High-dose fludrocortisone therapy was transiently required in a female neonate with 21-hydroxylase deficiency Abstract. For salt-wasting 21-hydroxylase deficiency (21OHD), fludrocortisone (FC) is usually supplemented at 0.05–0.2 mg/d dose. To date, no report has described 21OHD neonates requiring > 0.4 mg/d of FC. Our female 21OHD patient was lethargic and experienced weight loss with hyponatremia (133 mEq/L), hyperkalemia (6.5 mEq/L), and elevated active renin concentration (ARC, 1942.2 pg/mL) at 6 days of life. Hydrocortisone and FC replacement were initiated. FC dose was gradually increased to 0.4 mg/d at 21 days of life, but her hyperkalemia (6.4 mEq/L) and high ARC (372.3 pg/mL) persisted. We increased FC to 0.6 mg/d and used a low-potassium and high-sodium formula. Hyperkalemia subsequently improved. At 33 days of life, the ARC decreased to 0.6 pg/mL and FC dosage was gradually decreased. At 3 months of age, the low-potassium and high-sodium formula was discontinued, but the serum potassium level was normal and ARC remained low at 0.1 mg/d of FC. We speculated that severe mineralocorticoid resistance was the reason why her hyperkalemia persisted even with 0.4 mg/d of FC; however, the pathophysiology of transiently severe resistance to FC in this patient is unknown. In conclusion, 21OHD neonates may show severe salt-wasting that transiently require > 0.4 mg/d of FC. Introduction 21-hydroxylase deficiency (21OHD) is the most common cause of congenital adrenal hyperplasia. 21OHD is classified into severe salt-wasting, less severe simple-virilizing, and the least severe non-classic forms. Patients with salt-wasting 21OHD experience weight loss, poor feeding, dehydration, hyponatremia, and hyperkalemia owing to cortisol and aldosterone deficiencies usually during the second week of life (1,2). Treatment for salt-wasting 21OHD includes hydrocortisone (HC) and fludrocortisone (FC), potent glucocorticoids, and mineralocorticoids. FC is usually supplemented at 0.05-0.2 mg/d dose, and some affected neonates may require up to 0.4 mg/d FC (3). To date, no report has described affected neonates requiring > 0.4 mg/d of FC. Here, we report a neonate with 21OHD whose hyperkalemia was not normalized with 0.4 mg/d of FC administration. Case Report The clinical course of the patient is summarized in Fig. 1. The patient was the first child of healthy nonconsanguineous Japanese parents. The pregnancy was uncomplicated, and the mother delivered vaginally at 40 wk of gestation without asphyxia. The birth weight of the neonate was 3210 g (+ 0.64 standard deviation) and the length was 49.0 cm (−0.24 standard deviation). Clitoromegaly and labial fusion were observed on the first day after birth. She was referred to our hospital after six days of life. She was lethargic, and her weight had decreased by 96 g since the previous day. Blood examinations showed hyponatremia (133 mEq/L), hyperkalemia (6.5 mEq/L), elevated active renin concentration (ARC; 1942.2 pg/mL, reference for adults, 3.2-36.3), normal creatinine levels (0.42 mg/dL; reference for neonates of 6 days of life, 0.32-0.52), and elevated adrenocorticotropic hormone (1060 pg/mL; reference for adults, 7.2-63.3). We tentatively diagnosed the patient with adrenal crisis and initiated treatment with fluid replacement and continuous intravenous HC injection in saline at 100 mg·m -2 ·d -1 . Ultrasonography revealed swelling of the adrenal glands without lobulation. The filter paper blood 17α-hydroxyprogesterone concentration was 122.2 ng/mL. The urine steroid profile showed elevated pregnanetriolone (4.604 mg/gCr; age-specific cutoffs of 0-10 days of life, 0.06) and elevated ratios of 11β-hydroxyandrosterone/tetrahydroaldosterone and 11β-hydroxyandrosterone/pregnanediol (30.65 and 1.83, respectively; the cutoffs to distinguish between 21OHD and P450 oxidoreductase deficiency, 0.80 and 1.0, respectively) (4). Therefore, the patient was diagnosed with 21OHD. At 7 days of life, her poor sucking ability improved. At 9 days of life, the serum concentrations of sodium (137 mEq/L) and potassium (5.6 mEq/L) were normal and creatinine levels decreased (0.26 mg/dL). On the same day, intravenous administration of HC was discontinued and oral HC administration was changed to 60 mg·m -2 ·d -1 . At 11 days of life, she became lethargic and again showed poor sucking. Her weight had decreased by 400 g on the previous day. Blood tests revealed hyponatremia (130 mEq/L), hyperkalemia (7.7 mEq/L), and elevated ARC (1946.8 pg/mL). She had no signs of infection such as fever, apnea, vomiting, elevated C-reactive protein level, or high white blood cell count. The patient was diagnosed with a second adrenal crisis. Continuous intravenous administration of HC (100 mg·m −2 ·d −1 ) was resumed. At 14 days of life, hyponatremia (134 mEq/L) and hyperkalemia (5.7 mEq/L) persisted despite the oral administration of sodium polystyrene sulfonate. Therefore, oral administration of FC at 0.2 mg/d and sodium chloride at 1.0 g/d was initiated. Hyperkalemia (6.6 mEq/L) and high ARC (462.7 pg/mL) did not improve. At 18 days of life, FC was increased to 0.4 mg/d. At 21 days of life, she gained weight, but hyperkalemia (6.4 mEq/L) and high ARC (372.3 pg/mL) persisted. We increased FC to 0.6 mg/d and changed a regular formula to a low-potassium and high-sodium formula. Subsequently, hyponatremia and hyperkalemia improved. A blood test at 25 days of life showed that ARC tended to decline but was still high (206.2 pg/mL). During the course of treatment, the patient showed no significant changes in blood pressure or heart rate. At 33 days of life, ARC decreased to 0.6 pg/ mL. We started to gradually reduce FC dosage, which reached to 0.1 mg/d at 3 months of age. Hyponatremia and hyperkalemia did not recur when the low-potassium and high-sodium formulas were discontinued and ARC remained low (0.6 pg/mL). After 7 months, the patient had no recurrence of electrolyte abnormalities or adrenal crisis. Since birth, she had no symptoms suggestive of urinary tract infection. At 22 days of life, repeated ultrasonography showed no kidney or urethral malformation. Secondary pseudoaldosteronism was unlikely. After obtaining informed consent from her parents, we extracted the genomic DNA from peripheral blood samples of the patient and her parents. Sanger sequencing of CYP21A2 (NM_000500.9) suggested that: i) her mother had a heterozygous c.293-13C>G variant, ii) her father had on the same allele both c.1069C>T (p.Arg357Trp variant) and the deletion encompassing c.293-13C, and iii) the proband was compound heterozygous for alleles with pathogenic alterations, individually transmitted by her parents (Fig. 2). These two nucleotide substitutions are pathogenic variants (5). Exome sequencing of the proband revealed no pathogenic variants in the NR3C2, SCNN1A, SCNN1B, or SCNN1G genes associated with pseudohypoaldosteronism. Discussion In our patient with 21OHD, the high potassium levels were not controlled by administration of FC at 0.4 mg/d dose for 3 days but normalized with 0.6 mg/d FC and a low-potassium and high-sodium formula. We Clin Pediatr Endocrinol doi: 10.1297/cpe.31.2021-0066 considered that observation for 3 days was sufficient for evaluation of the treatment, as intravenous HC immediately shows mineralocorticoid activity. To the best of our knowledge, this is the first report of a neonate with 21OHD requiring > 0.4 mg/d of FC. Hyperkalemia at 6 days of life might, at least in part, be attributed to a low glomerular filtration rate because the serum potassium level was normalized with decreased serum creatinine. In addition, we speculated that severe mineralocorticoid resistance was the reason why her hyperkalemia persisted even at 0.4 mg/d of FC. During administration of FC (0.4 mg/d), her sodium intake, except for the regular formula, was 9.6 mmoL/ kg/d, including oral sodium chloride, intravenous fluid therapy, and sodium polystyrene sulfonate. A high dosage of sodium supplementation did not improve electrolyte imbalance in our patient (6). Her weight gain was good, and adrenocorticotropic hormone levels normalized with oral administration of 60 mg/m 2 /d HC. These findings suggest that there was no intestinal malabsorption. Congenital and secondary pseudohypoaldosteronisms were excluded because there were no urinary tract infections or malformations and no pathogenic variants in the NR3C2, SCNN1A, SCNN1B, and SCNN1G genes. Because the patient had relatively common pathogenic alterations in the CYP21A2 gene, the CYP21A2 genotype could not explain her atypical phenotype. During her clinical course, her severe mineralocorticoid resistance gradually improved after 33 days of life. In a previous study, the expression of mineralocorticoid receptors (MRs) in the distal nephron was low during the perinatal period and increased progressively after birth (7). In another study, hypovolemia induced dephosphorylation of MR in intercalated cells of the collecting duct, resulting in suppression of potassium excretion by aldosterone (8). Our patient may still have hypovolemia at 0.2 mg/d of FC because of no weight gain and high ARC. The effects of hypovolemia persisted for several days. Although the exact mechanism of mineralocorticoid resistance remains unclear, the combination of physiological and pathological compromises in MR expression can be related to transiently severe mineralocorticoid resistance in our patient. We were unable to determine the cause of the adrenal crisis at 11 days. We speculate that reducing the dosage of HC to 60 mg · m −2 · d −1 , changing from continuous intravenous infusion to oral administration, or both precipitated a second adrenal crisis. A previous study showed that serum cortisol concentrations widely fluctuated upon oral administration of HC but remained stable upon continuous intravenous infusion (9). This implied that in our patient, unstable glucocorticoid absorption upon oral administration could have resulted in unstable serum cortisol concentrations, which may have triggered adrenal crisis. Our patient had a clinical course similar to that of most 21OHD patients who developed clinical symptoms and metabolic imbalances, including hyponatremia and hyperkalemia, 1-2 weeks after birth (3,10). Thus, in newborns with suspected 21OHD, electrolyte imbalances and clinical symptoms should be closely monitored during the first two weeks of life. Notably, our patient had only one atypical laboratory finding of highly elevated ARC before treatment initiation and 6 days after birth. A previous study examining 14 patients with 21OHD showed that the ARC concentration was 30-100 pg/mL in most patients and 150-160 pg/mL in a few patients (11). We speculate that extremely high ARC levels might be an indicator of the requirement for high-dose FC. Conclusion We report the case of a female neonate with 21OHD who showed severe salt-wasting and required > 0.4 mg/d of FC. The pathophysiology of the transient severe resistance to FC in this patient is unknown. Neonates with 21OHD may show severe salt-wasting and require > 0.4 mg/d of FC. Even during the treatment of HC and FC, electrolyte imbalance in 21OHD may occur or worsen 1-2 weeks after birth. Ethical statement: This study complied with all the relevant national regulations and institutional policies, was in accordance with the tenets of the Helsinki Declaration, and was approved by the Institutional Review Board at Keio University School of Medicine (Institutional Review Board number 20150104 and 20170130). Written informed consent was obtained from the patient's parents. Conflict of interests: The funding organizations played no role in the study design, collection, analysis, and interpretation of data, in the writing of the report, or in the decision to submit the report for publication. Clin Pediatr Endocrinol Increased Water Content in Periventricular Caps in Patients without Acute Hydrocephalus BACKGROUND AND PURPOSE: Periventricular caps are a common finding on MR imaging and are believed to reflect focally increased interstitial water content due to dysfunctional transependymal transportation rather than ischemic-gliotic changes. We compared the quantitative water content of periventricular caps and microvascular white matter lesions, hypothesizing that periventricular caps associated with increased interstitial fluid content display higher water content than white matter lesions and are therefore differentiable from microvascular white matter lesions by measurement of the water content. MATERIALS AND METHODS: In a prospective study, we compared the water content of periventricular caps and white matter lesions in 50 patients using a quantitative multiple-echo, gradient-echo MR imaging water-mapping sequence. RESULTS: The water content of periventricular caps was significantly higher than that of white matter lesions (P = .002). Compared with normal white matter, the mean water content of periventricular caps was 17% ± 5% higher and the mean water content of white matter lesions was 11% ± 4% higher. Receiver operating characteristic analysis revealed that areas in which water content was 15% higher compared with normal white matter correspond to periventricular caps rather than white matter lesions, with a specificity of 93% and

a sensitivity of 60% (P < .001). There was no significant correlation between the water content of periventricular caps and whole-brain volume (P = .275), white matter volume (P = .243), gray matter volume (P = .548), lateral ventricle volume (P = .800), white matter lesion volume (P = .081), periventricular cap volume (P = .081), and age (P = .224). CONCLUSIONS: Quantitative MR imaging allows differentiation between periventricular caps and white matter lesions. Water content quantification of T2-hyperintense lesions may be a useful additional tool for the characterization and differentiation of T2-hyperintense diseases. RESULTS: The water content of periventricular caps was significantly higher than that of white matter lesions (P ϭ .002). Compared with normal white matter, the mean water content of periventricular caps was 17% Ϯ 5% higher and the mean water content of white matter lesions was 11% Ϯ 4% higher. Receiver operating characteristic analysis revealed that areas in which water content was 15% higher compared with normal white matter correspond to periventricular caps rather than white matter lesions, with a specificity of 93% and a sensitivity of 60% (P Ͻ .001). There was no significant correlation between the water content of periventricular caps and whole-brain volume (P ϭ .275), white matter volume (P ϭ .243), gray matter volume (P ϭ .548), lateral ventricle volume (P ϭ .800), white matter lesion volume (P ϭ .081), periventricular cap volume (P ϭ .081), and age (P ϭ .224). CONCLUSIONS: Quantitative MR imaging allows differentiation between periventricular caps and white matter lesions. Water content quantification of T2-hyperintense lesions may be a useful additional tool for the characterization and differentiation of T2-hyperintense diseases. ABBREVIATIONS: iNPH ϭ idiopathic normal pressure hydrocephalus; PVC ϭ periventricular cap; WML ϭ white matter lesions P eriventricular caps (PVCs), smooth T2-hyperintense areas around the lateral ventricles (mostly the frontal and posterior horns, Fig 1), are a common finding on cranial MR imaging, especially in elderly patients. Pathologic studies have shown that PVCs do not correspond to ischemic-gliotic changes, but rather to an area of finely textured myelin associated with denudation of the ventricular ependymal lining. 1 Because the ependyma provides a bidirectional barrier and transport system for CSF and interstitial fluid exchange, age-related ependymal denudation might impair normal transportation of interstitial fluid into the ventricles, leading to an accumulation of interstitial fluid, which converges from the surrounding white matter, around the frontal horns of both lateral ventricles. 1 According to this theory, the increased T2 signal intensity in PVCs would reflect, at least in part, edematous rather than pure gliotic white matter changes. However, gliotic white matter changes caused by microvascular disease are also a frequent occurrence in the general population. 2 Because conventional T2-weighted sequences do not allow one to differentiate between those 2 types of lesions, alternative imaging sequences are necessary to differentiate these 2 pathologies. Quantitative MR imaging allows the quantification of water content of brain lesions within a clinically acceptable acquisition time of a few minutes. 3,4 For instance, it has been shown that quantitative water mapping allows one to detect an increase in cerebral water content in hepatic encephalopathy and to evaluate the surrounding edema in brain tumors. 5,6 We hypothesized that quantitative MR imaging allows differentiation between PVCs and microvascular white matter lesions (WML). Hence, the aim of this study was to investigate whether PVCs, representing interstitial fluid due to dysfunctional transependymal transportation, contain an increased water content compared with gliotic microvascular deep white matter lesions, using a quantitative water-mapping sequence. MATERIALS AND METHODS After approval from our local ethics board (faculty of medicine, RWTH Aachen University), we prospectively scanned all patients who presented with neurovascular symptoms between May 2014 and February 2017 and who agreed to participate in our study (n ϭ 177). For this analysis, we included all patients who had both PVCs and WML with a volume of at least .1 cm 3 . This cutoff was chosen to address lesions that are clearly assessable on clinical MR imaging with a typical slice thickness of 3-5 mm, thereby reducing partial volume effects on our measurements. We excluded patients with acute or subacute stroke adjacent to the investigated areas and patients with pathologies other than PVCs and WML, such as brain tumors or inflammation. This exclusion left 50 patients included in our study. No patient had clinical or radiologic signs of acute hydrocephalus or idiopathic normal pressure hydrocephalus (iNPH). Analysis To delineate PVCs and WML, we used a semiautomatic segmentation approach to allow objective segmentation. Initially, we used the open-source deep learning framework DeepMedic (Version 0.6.1; https://biomedia.doc.ic.ac.uk/software/deepmedic/), which is used for automated segmentation, to segment the hyperintensities in the T2-FLAIR sequence. 7 We trained the neural network with 30 cases of voxelwise manually annotated ROIs as "PVCs" and "WML," depending on their morphology and location. We defined PVCs as smooth periventricular caps at the frontal and posterior horns of the lateral ventricles (grade 1 PVCs according to Thomas et al 8 ), whereas irregular T2-hyperintense periventricular regions were not regarded as PVCs. WML were defined as patchy or confluent T2-hyperintense areas in the deep cerebral white matter. Areas where we could not clearly distinguish PVCs and WML were excluded altogether from our analysis. For our final analysis, 2 independent observers examined all segmented areas and adjusted the ROIs manually in a consensus reading, for example, to modify ROIs contiguous with the ventricles that could have been affected by partial volume effects. The final ROIs were then used in the water-mapping acquisitions to measure the quantitative water content of PVCs and WML. In addition, we investigated whether there was a gradient of water content from central periventricular white matter to more peripheral subcortical white matter by measuring 3 ROIs in the periventricular white matter, the centrum semiovale, and the subcortical white matter. Thus, we manually placed 3 respective ROIs in normal white matter on 1 axial slice adjacent to the examined T2 hyperintensities. Statistical Analysis After testing for data distribution with a Shapiro-Wilk test, we performed a paired Student t test to determine whether the water content between areas of PVCs and WML differed significantly. We conducted an ANOVA to assess whether there was a gradient between the central and peripheral water content. We performed a receiver operating characteristic analysis to determine the sensitivity and specificity of water content measurements for PVCs and WML distinction. We examined correlations between PVC water content and brain volume, lateral ventricle volume, lesion volume, and age using the Pearson correlation coefficient. Continuous parametric variables are presented as mean Ϯ SD, and P values with an ␣ level Ͻ.05 were considered statistically signifi- cant. All statistical analyses were calculated using the SPSS software package, Version 25 (IBM, Armonk, New York). The absolute water content and the relative increase of water content of PVCs was significantly higher than that of WML (P ϭ .002). The mean absolute water content in PVCs was 81% Ϯ 2% (median, 81%; range, 77%-85%). The water content in PVCs was 17% Ϯ 5% (median, 16%; range, 7%-26%) higher than that of normal white matter (Fig 2). The mean absolute water content in WML was 77% Ϯ 2% (median, 77%; range, 71%-81%). The water content in WML was 11% Ϯ 4% (median, 11%; range, 5%-19%) higher than that of normal white matter (Fig 2). The water content of PVCs was higher than that of WML in all patients. Lesion volume had no significant impact on water content in PVCs and WML (P ϭ .953). Receiver operating characteristic analysis revealed that a 15% higher water content corresponded to PVCs rather than WML with a specificity of 93% and a sensitivity of 60% (P Ͻ .001, area under the curve ϭ .842). DISCUSSION The microscopic nature of PVCs in patients without acute hydrocephalus is not fully understood. The literature suggests that PVCs might be secondary to dysfunctional transependymal transportation of interstitial fluid into the ventricles; thus, accordingly, it is expected that PVCs show higher water content compared with gliotic deep white matter lesions. 1 In our study, we found that the water content of PVCs was indeed significantly higher than that of deep WML. Our receiver operating characteristic analysis revealed that a lesion with a 15% or higher water content compared with normal white matter corresponds to PVCs rather than deep WML, with an acceptable sensitivity and specificity. There was no significant correlation between the water content of PVCs and brain volume, lateral ventricle volume, lesion volume, and age. Because PVCs are located around the ventricles and WML are located peripherally, we investigated whether the water content difference was due to a water content gradient in healthy white matter with high water content in the periventricular areas and low content in the periphery. We assessed the water content in the respective regions and did not detect any such gradient, suggesting that the high water content in PVCs is, in fact, evidence of a histopathologic difference between deep WML and PVCs. Our study does not allow determining the exact pathophysiology of PVCs, in particular because our study lacks longitudinal imaging that could elucidate the etiology of PVCs. Hence, water content serves as a surrogate imaging marker that confirms pathologic studies that showed that there is increased water content accumulating from white matter surroundings or the intense venous network in this region. 9,10 This work, therefore, serves as an in vivo validation of previous pathologic studies. The results of this study may be translated to patients with iNPH and may help understand the pathophysiology of that disease: For instance, it is still unclear whether periventricular changes in patients with iNPH represent transependymal edema or ischemic gliosis. In fact, it has been suggested that iNPH is primarily driven by gliotic white matter infarction, which diminishes brain tissue compliance, especially because iNPH is associated with vascular risk factors such as diabetes and arterial hypertension. [11][12][13] On the other hand, abnormalities of CSF circulation are believed to be the primary cause of iNPH, [14][15][16] supported by the fact that clinical symptoms and the degree of white matter lesions improve in some patients after ventricular shunting, the latter possibly depending on whether these lesions were gliotic or edematous. [17][18][19] Theoretically, quantitative water mapping could be used in these patients to distinguish periventricular glioticischemic from edematous white matter changes in an attempt to elucidate the underlying nature of these changes and eventually predict shunt-responsiveness. Limitations A major limitation of our study is the lack of an in vivo criterion standard for the distinction between gliotic and edematous T2 hyperintensities, which leads, to some extent, to a speculative interpretation of our data. In addition, it is not evident whether the presumably gliotic WML in our included patients are, in fact, caused by ischemia. To address this issue, we included only patients in whom microvascular ischemic gliosis was the most probable cause for their WML. Due to the study design, our study population consists solely of patients with neurovascular symptoms; inclusion of healthy, asymptomatic controls or patients presenting a wider range of symptoms would allow a better generalization of the findings. A strength of our study is the inclusion of patients with both PVCs and WML, which allows an intraindividual analysis. The fact that the water content of PVCs surpassed the water content of WML in every single patient underlines that averaging effects had no major impact on our results. CONCLUSIONS We found that the water content of periventricular caps was significantly higher than that of deep white matter lesions; the latter were most probably of microvascular origin. Our results show that a lesion with a 15% or higher water content compared with normal white matter corresponds to PVCs rather than WML. Inversely, this finding also implies that it is possible to use quantitative water mapping to distinguish gliotic white matter lesions from other T2-hyperintense lesions. Quantifying the water content of periventricular T2-hyperintense white matter changes may be a useful additional tool for the differentiation of T2 hyperintensities in iNPH or other diseases such as multiple sclerosis. Alterations of default mode functional connectivity in individuals with end-stage renal disease and mild cognitive impairment Background Mild cognitive impairment (MCI) occurs frequently in many end stage renal disease (ESRD) patients, may significantly worsen survival odds and prognosis. However, the exact neuropathological mechanisms of MCI combined with ESRD are not fully clear. This study

examined functional connectivity (FC) alterations of the default-mode network (DMN) in individuals with ESRD and MCI. Methods Twenty–four individuals with ESRD identified as MCI patients were included in this study; of these, 19 and 5 underwent hemodialysis (HD) and peritoneal dialysis (PD), respectively. Another group of 25 age-, sex- and education level-matched subjects were recruited as the control group. All participants underwent resting-state functional MRI and neuropsychological tests; the ESRD group underwent additional laboratory testing. Independent component analysis (ICA) was used for DMN characterization. With functional connectivity maps of the DMN derived individually, group comparison was performed with voxel-wise independent samples t-test, and connectivity changes were correlated with neuropsychological and clinical variables. Results Compared with the control group, significantly decreased functional connectivity of the DMN was observed in the posterior cingulate cortex (PCC) and precuneus (Pcu), as well as in the medial prefrontal cortex (MPFC) in the ESRD group. Functional connectivity reductions in the MPFC and PCC/Pcu were positively correlated with hemoglobin levels. In addition, functional connectivity reduction in the MPFC showed positive correlation with Montreal Cognitive Assessment (MoCA) score. Conclusion Decreased functional connectivity in the DMN may be associated with neuropathological mechanisms involved in ESRD and MCI. Background End stage renal disease (ESRD) correspond to stage 5 chronic kidney disease (CKD), when chronic renal failure has progressed to the point that the kidneys are permanently functioning at less than 10% of normal capacity [1]. In additional to renal failure, mild cognitive impairment (MCI) is a common comorbidity in ESRD cases [2]. This cognitive problem in ESRD could be caused by uremia, thiamine deficiency, hypertension, dialysis, transplant rejection and/or electrolyte disturbances [3]. Patients with MCI are of high risk of developing dementia, which may significantly worsen survival odds and prognosis [4]. However, the exact neuropathological mechanisms of MCI combined with ESRD remain unclear, thus hampering the development of efficient treatments. Conventional medical imaging has played an important role in exploring the structural and functional brain mechanisms associated with ESRD, such as cerebral infarction, intracranial hemorrhage, subcortical white matter lesions, volume reduction, and metabolic disturbances [5][6][7]. However, due to the limitation of insensitivity to early morphological changes, existing imaging approaches in the clinical setting may be improved by additional assessment of functional networks. Recently, much attention has been paid to resting state functional magnetic resonance imaging (rsfMRI) [8,9], which constitutes a novel paradigm that examines spontaneous brain function by using blood oxygen leveldependent contrast in the absence of a task. Changes of blood oxygen saturation in some brain regions indicate functional alterations in these areas. Therefore, changes of functional connectivity in rsfMRI could reflect the functional status of the brain. Alterations in resting state networks have been identified in many diseases, such as Alzheimer's disease [10], traumatic brain injury [11], hepatic encephalopathy [12], and ESRD [13][14][15][16]. In these studies, abnormal functional connectivity of the default mode network (DMN) has attracted great attention. The DMN is one of the resting state functional networks, which contain meaningful information related to spontaneous brain activity [10] and may be affected by neurological complications related to ESRD. Luo et al. [15] reported that ESRD patients show lower ALFF values in DMN regions compared with healthy control subjects. Ma et al. [14] and Qiu et al. [16] found that ESRD patients exhibit significantly decreased functional connectivity in several regions of the DMN by seedbased analysis. In contrast to seed-based analysis, which heavily depends on a priori seeds defined, independent component analysis (ICA) requires no prior knowledge of a network, and hence may provide complementary/confirmatory information [9]. Using ICA, Ni et al. [13] extracted the DMN component in ESRD patients successfully, and found that the patients exhibit significantly decreased functional connectivity in the PCC, precuneus, and MPFC compared with control subjects, with functional connectivity of the MPFC positively correlated with digital symbol test score. These findings of DMN alterations associated with ESRD are insightful; however, previous studies could not perform elaborate grouping according to the degree of cognitive impairment, and did not rule out the effects of comorbidities, such as diabetes and cardiovascular disease, in conclusions made about ESRD. In the present study, we specifically excluded ESRD patients complicated with diabetes mellitus, and used ICA to investigate FC alterations of the DMN. Focusing on ESRD patients complicated with MCI, we also examined potential associations of FC changes of the DMN with neuropsychological and clinical variables. Subjects This prospective study was approved by the Medical Research Ethics Committee of our hospital, and written informed consent from each subject was obtained before the study. From April 2016 to June 2017, 31 individuals (including outpatients and inpatients; all right-handed) with ESRD combined with MCI were initially recruited in this study. For inpatients, the cause of hospitalization were concurrent diseases, such as lung infection and hypertension, which may not impact the functional status of the brain. Primary diseases in the ESRD group included chronic glomerulonephritis (n = 7), polycystic kidney (n = 4), IgA nephropathy (n = 2), nephrotic syndrome (n = 2), hypertension (n = 1), and unknown ailments (n = 8). MCI was diagnosed according to Petersen criteria [17] for amnestic MCI: 1) memory complaint corroborated by an informant, 2) objective memory impairment for age, education and gender, 3) essentially preserved general cognitive function, 4) largely intact functional activities, 5) no dementia. For criterion 2), the Clinical Dementia Rating scale (CDR) [18] was used to identify MCI. CDR scores were obtained via a semistructured interview of patients and informants, and the subject's cognitive status was rated in six domains of functioning, including memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care. Performances in these domains were assessed by a clinician with 9 years of experience in nephrology (HZ). When a subject had an overall CDR index of 0.5, the diagnosis of MCI was established. All subjects completed laboratory tests (serum creatinine, blood urea nitrogen and hemoglobin) within 72 h before MRI for evaluating renal function and the physical status. Dialysis modality and duration were obtained from the patient's medical history. Neuropsychological tests, including Mini-Mental State Examination (MMSE) [19] and MoCA [20], were also completed within one hour before MR imaging to quantify objective memory impairment and essentially preserved general cognitive function of ESRD cases. Exclusion criteria were: (a) a history of psychiatric disorders; (b) major neurologic disorders such as severe traumatic brain injury, stroke, epilepsy and tumor); (c) drug or alcohol abuse; (d) a history of diabetes; and (e) head motion more than 3.0 mm or 3.0°during fMRI. Based on the above criteria, 7 participants were excluded for cerebral hematoma (n = 2), a history of traumatic injury (n = 1) and excessive head motion (n = 4), resulting in a study sample of 24 participants (11 men and 13 women; mean age of 50.58 ± 9.49, ranging from 31 to 65 years). In this sample, 19 and 5 participants underwent hemodialysis (HD) and peritoneal dialysis (PD), respectively. Twenty-five matching controls in terms of age, sex ratio and education level were recruited from the local community. Self-report of good health with normal renal function, confirmed by laboratory data from community doctors within 12 months of testing, no history of psychiatric or neurologic disease, and normal-abdominal ultrasonographic (US) imaging. All participants were right-handed. The control group completed the same procedure of MRI and neuropsychological tests within one hour before MRI imaging. MR imaging data acquisition A 3.0 Tesla MR scanner (Achieva, Philips, Best, the Netherlands) and a 16-channel phased array head coil were used to acquire all MRI data. During MRI scans, participants were instructed to lie quietly with eyes closed but staying awake. A foam pad was used to reduce head motion. The degree of patient cooperation was verified after the examination. Conventional imaging sequences, including axial T1-and sagittal T1-weighted images and coronal T2 fluid-attenuated inversionrecovery images, were obtained for every subject to detect clinically silent lesions. For co-registration and normalization of resting-state functional MR imaging data, high-resolution T1-weighted 3D anatomical images were obtained in sagittal slices using a magnetizationprepared rapid gradient-echo sequence (TR/TE, 8.0/3.9 ms; slice thickness, 1 mm; 160 slices; flip angle, 8°; field of view, 250 × 250 mm 2 ; acquisition matrix, 252 × 227; voxel size in acquisition, 0.99 × 1.10 × 2.00 mm 3 ; voxel size after reconstruction, 0.98 × 0.98 × 1.00 mm 3 ). Functional data were collected by a gradient-echo echoplanar sequence in axial slices parallel to the line through the anterior and posterior commissures: TR/TE, 2000/30 ms; slice thickness, 4 mm; gap between slices, 0.4 mm; field of view, 240 × 240 mm 2 ; acquisition matrix, 120 × 117; resolution, 2.00 × 2.05 × 4.00 mm 3 ; flip angle, 90°; fold-over direction, AP; EPI factor, 41. Each rsfMRI sequence contained 250 volumes with 30 slices. Data preprocessing Data were preprocessed with statistical parametric mapping (SPM8, http://www.fil.ion.ucl.ac.uk/spm/) in Matlab (MathWorks, Natick, Mass). First, 3D anatomical data were reoriented, skull stripped, and segmented into gray matter, white matter and cerebrospinal fluid regions. Then, the first 10 volumes of rsfMRI data were discarded for magnetization equilibrium and participant adaptation. Slice timing and head motion corrections were performed subsequently followed by T1-EPI coregistration. Afterwards, spatial normalization was performed based on the unified segmentation to the T1 image and the fMRI data was warped into the standard space of the Montreal Neurologic Institute (MNI) template with a resampled voxel size of 3 × 3 × 3 mm 3 . Finally, images were spatially smoothed by convolution with an isotropic Gaussian kernel of FWHM = 8 mm. Independent component analysis Group ICA was performed with the software package of GIFT (Vision 2.0e; http://icatb.sourceforge.net/). GIFT linearly decomposes 4D rsfMRI data into spatially independent sources, each of which is considered an intrinsically connected brain network [21]. To determine the number of independent components, dimension estimation was carried out individually using the minimum description length criterion [22], and averagely 21 components were obtained. This averaged independent component was used for each subject for independent component analysis separation. Then, functional MR imaging data of all subjects in both groups were concatenated in the temporal dimension, and the data set was separated into independent components by the InfoMax algorithm. After the ICA, a DMN template described in previous study [23] was used to select the best fit of the remaining low-frequency components in each subject. The template-matching procedure was realized by subtracting the average IC z score of voxels in the template from that outside the template, and the component with the maximum difference (goodness of fit) was selected [13]. The z scores used here reflected the degree to which a given voxel's time series correlated with the time series corresponding to a specific independent component, scaled by the standard deviation of the error term. Therefore, the z score could be used to reflect the amplitude of regional activity relative to the background noise. In this study, component 3 had the maximum difference (0.323), but followed closely by component 15 (0.274), so both were selected for subsequent group comparisons. The IC map of component 3 predominantly consisted of the posterior cingulate cortex (PCC), precuneus, bilateral inferior parietal and medial prefrontal cortex (MPFC), which was similar to a posterior-DMN. The IC map of component 15 was predominantly composed of the MPFC, PCC, bilateral inferior parietal, which was similar to an anterior-DMN [24]. Statistical analysis With SPM8, statistical analysis for rsfMRI data was performed by one sample (within group) and independentsamples (group contrast) t-tests. A voxel-wise threshold of P < 0.001 was applied, and false positives from multiple comparisons were controlled by the Alphasim procedure (Alphasim; http://afni.nih.gov/afni/docpdf/AlphaSim.pdf), with a cluster size greater than 22 voxels to remove falsepositives and maintain true-positive sensitivity. For each component, within-group clusters of control and ESRD patients were combined with a logical "AND" operation, and subsequent group contrast was only considered in this conjunction mask. In two-sample t-tests, age and sex were included as covariates, and the voxel-wise threshold was set at P < 0.05. The result was corrected for multiple comparison with cluster sizes greater than 428 voxels (component 3) and 325 voxels (component 15). These cluster thresholds were determined by a Monte Carlo simulation implemented in the DPABI software package (http://rfmri. org/DPABI) with parameters of P < 0.05/voxel and 3D Gaussian smoothness of FWHM = 8

mm [25]. Statistical analysis for demographic and clinical data was performed by SPSS (version 16.0; SPSS, Chicago, III). Significant regions in group differences of components 3 and 15 were considered regions of interest (ROIs) [26] for individual IC z scores to be extracted from. Pearson correlation was used to assess associations of average IC z score of each participant in the ROI with clinical variables/psychological parameters (serum creatinine, blood urea nitrogen, hemoglobin, MMSE and MoCA scores) in the patient group. Reproducibility analysis A split-half analysis was used to assess result reproducibility. The control and ESRD groups were both divided into two age-and gender-matched subgroups: control-1 (13 participants; 5M8F; age, 46.23 ± 12.73 years), control-2 (12 participants; 4M8F; age, 46.42 ± 11.74 years), ESRD-1 (12 participants; 5M7F; age, 49.92 ± 9.95 years) and ESRD-2 (12 participants; 6M6F; age, 51.17 ± 9.59 years). For each significant region identified in whole-sample group comparison, we extracted IC z scores from the ROIs and performed group comparison again across different subgroups. If differences between subgroups of the same diagnosis (e.g. control-1 vs. control-2) were insignificant but those between subgroups of different diagnoses (e.g. control-1 vs. ESRD-1) significant, the whole-sample results were considered to be reproducible. Results The demographic and clinical data of all participants are summarized in Table 1. While group differences in age (P = 0.173), gender (P = 0.484), and education level (P = 0.386) were non-significant, the ESRD group had significantly lower MMSE (P = 0.01) and MoCA scores (P < 0.001). The duration of dialysis for individuals with ESRD was more than 3 months. Blood urea nitrogen and serum creatinine levels were out of respective normal ranges. Group comparisons of FC of the DMN are graphically shown in Fig. 1 (component 3) and Fig. 2 (component 15). Correlation analysis revealed that average IC z score in the MPFC was positively correlated with MoCA score (r = 0.409, P = 0.047) and hemoglobin levels (r = 0.467, P = 0.021) (Fig. 3). Meanwhile, average IC z score in PCC/Pcu had a positive correlation with hemoglobin levels (r = 0.619, P = 0.001) as well (Fig. 3). Reproducibility data for evaluating average IC z scores of PCC/Pcu and the MPFC among subgroups are shown on Fig. 4. There were no significant differences in PCC/ Pcu or MPFC results between control-1 and control-2, as well as between ESRD-1 and ESRD-2. Meanwhile, differences between control (control-1, control-2) and ESRD (ESRD-1, ESRD-2) subgroups in PCC/Pcu and Discussion This study demonstrated that impairment of DMN functional connectivity in individuals with ESRD and MCI at rsfMRI. Individuals with ESRD and MCI showed decreased functional connectivity in the PCC, precuneus and MPFC regions. Furthermore, functional connectivity reduction in the MPFC was positively correlated with MoCA score and hemoglobin levels in individuals with ESRD and MCI. Functional connectivity decreases in the PCC and precuneus were positively correlated with hemoglobin levels as well. The present findings complemented previous studies [5,27,28] suggesting that metabolic and structural abnormalities are associated with ESRD. In this study, DMN regions with impaired functional connectivity in individuals with ESRD and MCI, which comprised the anterior hub (MPFC) and posterior hub (PCC) of the DMN, were different from those of MCI caused by degeneration or small vessel disease (SVD). [29]. Deactivation in PCC/Pcu and the MTL is mainly considered a consequence of amyloid accumulation and tau pathology [30]. DMN changes in individuals with SVD are mainly found in the anterior-DMN, mainly the MPFC [31]; the possible mechanism is that patients with SVD have disconnected white matter tracts in the frontal region [31]. The mechanism of impaired FC in individuals with ESRD is likely a combined action of multiple factors, including creatinine and urea accumulation, SVD associated with kidney failure, and dialysis. Posterior leukoencephalopathy might be a promoting factor. Indeed, posterior leukoencephalopathy is not uncommon in ESRD patients undergoing dialysis [32], and the PCC as the posterior DMN hub is often involved in the complications of impaired cerebrovascular autoregulation, endothelial injury and elevated plasma concentrations of natriuretic peptides. Brain microvascular endothelial dysfunction induced by uremic toxins can directly result in SVD; dialysis may bring about the decrease of plasma osmotic pressure, and then induce brain edema; posterior leukoencephalopathy can cause nerve cell dysfunction and even apoptosis, these factors may eventually lead to widespread DMN alteration mainly in the MPFC and PCC/Pcu. These changes of DMN on rsfMRI can be used as imaging markers for the diagnosis of ESRD patients accompanied with MCI in its early stage. We also found that functional connectivity decrease in the MPFC had a positive correlation with MoCA score, while functional connectivity reductions in the MPFC and PCC/Pcu were positively correlated with hemoglobin levels. MPFC is involved in cognitive processing related to self-awareness, episodic memory, and interactive modulation between internal brain activities and external tasks [33]. Impaired cognitive function might be attributed to functional connectivity alteration in the MPFC. Meanwhile, no significant correlation was found between FC in the MPFC and the MMSE score, which might be due to the weak sensibility of MMSE in evaluating cognitive impairment [14]. Positive associations of decreased functional connectivity in the MPFC and PCC/Pcu with hemoglobin levels were also observed. Previous studies have shown that anemia is associated with cognitive dysfunction in patients with ESRD [34]. This indicates that ESRD patients with MCI may benefit from enhanced hemoglobin levels. In a similar study, Ni et al. [13] also reported that reduced functional connectivity of the DMN is negatively correlated with serum creatinine levels, while we did not observe a significant correlation between FC and serum creatinine levels. We speculate that this discrepancy could be due to dialysis. Indeed, all patients in this study underwent dialysis; thus, serum creatinine levels were well managed. In contrast, participants in Ni et al. were not all being treated by dialysis. The limitations of this study should be mentioned. First, the sample size was relatively small, which may have affected the statistical power. Secondly, we did not distinguish between various dialysis modalities; serum potassium, hemoglobin, serum albumin, and brain natriuretic peptide levels differed between the HD and PD groups, which might lead to varying degrees of FC impairment in DMN regions. Finally, due to the crosssectional design, it was impossible to examine the dynamic changes of functional connectivity with ESRD progression, or even FC differences before and after dialysis. Therefore, a longitudinal study is warranted. Conclusion Aiming to improve previous studies focusing on DMN alterations associated with ESRD, we controlled confounding factors better with elaborate grouping and more rigorous exclusion criteria. The present findings indicated that FC in the DMN is impaired in individuals with ESRD and MCI, and such FC changes are correlated with MoCA score and hematocrit levels. These results complemented and/or extended previous reports with additional insights into potential neural underpinnings of asymptomatic ESRD combined with MCI. Pharmaceuticals and Environment: a web-based decision support for considering environmental aspects of medicines in use Purpose The database Pharmaceuticals and Environment is a non-commercial, freely available web-based decision support presenting compiled environmental information for pharmaceutical substances. It was developed by Region Stockholm and launched in 2016 at janusinfo.se. The purpose of this paper is to present the database, report on its current use, and reflect on lessons learned from developing and managing the database. Methods A standard operating procedure describes the work and content of the database, e.g., how information is retrieved, processed, and presented. Google Analytics was used for metrics. Issues related to the database have been discussed and handled by a reference group. The experiences from this work are presented. Results The database contains environmental hazard and risk information, primarily gathered from regulatory authorities and pharmaceutical companies. There are also assessments comparing substances within some groups of pharmaceuticals. The database is used by the Swedish Drug and Therapeutics Committees to include environmental aspects when recommending pharmaceuticals for health care providers. Page views show that users primarily look for information on commonly used substances, e.g., diclofenac and paracetamol/acetaminophen. Major problems for the development of the database are lack of data, lack of transparency, and discrepancies in the available environmental information. Conclusion In the absence of an adequate decision support produced by the regulatory authorities, we find the database Pharmaceuticals and Environment to be useful for Swedish Drug and Therapeutics Committees and health care providers, and it is our belief that the information can be valuable also in other settings. Introduction The use of pharmaceuticals is likely to continue to increase due to a growing aging population with chronic diseases, and access to new pharmaceuticals [1][2][3]. This is positive in terms of improved health for the treated individuals but at the same time poses an increased risk to the environment. Studies have already shown that chronic exposure to low concentrations of ethinylestradiol leads to feminization of male fish and that levonorgestrel can bioconcentrate in fish and disrupt oogenesis in frogs [4][5][6]. The massive decline in vulture populations in the Indian subcontinent in the 1990s was caused by diclofenac poisoning [7], and behavioral changes in fish have been shown in studies with oxazepam and citalopram [8,9]. In addition, antimicrobial resistance is an increasing problem and poses a serious threat to global health and antibiotic residues in the environment can be a contributing factor to this problem [10,11]. Furthermore, high levels of pharmaceuticals in the effluent from the manufacturing process can cause local environmental problems [12]. Region Stockholm, the provider of most of the health care in Stockholm county, has been committed to reduce the environmental impact of pharmaceuticals since the beginning of the 2000s [13]. This work has included analyses of water samples for active pharmaceutical ingredients (APIs) and information to the public to increase the awareness of environmental effects of pharmaceuticals. To be able to focus management measures on the most harmful substances, environmental information on APIs was needed and therefore the work on gathering information began in 2001 [13]. The environmental information, which was originally presented in a printed brochure and on a webpage, is now available in the public database Pharmaceuticals and Environment at janusinfo.se in a Swedish and an English version [14]. The aim of the database is to present information about environmental hazard and risk associated with APIs for human use on the Swedish market. In order to be an adequate tool for risk assessors and decision-makers, a database needs to provide accurate, updated, and easily accessible information that is adapted for the intended users [15]. There are other databases with the purpose of communicating research results and environmental information, thereby facilitating decision-making. LIF, the trade association for the research-based pharmaceutical industry in Sweden, presents environmental information for some of the medicinal products on the Swedish market at fass.se [16]. The European Medicines Agency (EMA) provides environmental risk assessments for medicinal products approved after 2006 in the European Public Assessment Reports (EPARs) at EMA's website [17]. The WikiPharma database and the NORMAN Ecotoxicology Database both provide ecotoxicity data [15,18]. The WikiPharma database was one of the outcomes of the Swedish MistraPharma research program during 2008-2015 which generated new data, e.g., identified pharmaceutical substances that pose risks in the aquatic environment, and issued recommendations for improved environmental risk assessments of human pharmaceuticals [19]. The ETOX database from the German Environmental Agency provides ecotoxicology data as well as information on various national and international environmental quality guidelines and limit values [20]. In the iPiE database, pharmaceutical companies present information on properties, environmental fate characteristics, and ecotoxicity of pharmaceuticals [21]. The US Environmental Protection Agency has developed the database ECOTOX which also contains data from ecotoxicity studies for pharmaceuticals [22]. The number of databases indicates that there is a need for decision support in this field, and that users may have different needs, hence the variety in focus and structure of the databases. The purpose of this paper is to present the database Pharmaceuticals and Environment, report on its current use, and reflect on lessons learned from developing and managing a database intended as decision support. Method Developing the database In Sweden, classification of environmental hazard of pharmaceuticals, presented per API, was initiated in 2001 by Region Stockholm and Apoteket AB (public owner of all pharmacies in Sweden at that time) [13]. In 2005, the work was jointly developed to include an environmental risk assessment by Region Stockholm, Apoteket AB, the

Swedish Medical Products Agency, LIF, and the Swedish Association of Local Authorities and Regions [13]. The classification focused on environmental hazard and risk associated with APIs for human use on the Swedish market. This work resulted in the environmental information presented per medicinal product at fass.se. However, having the information presented per product was not suitable for the needs of Region Stockholm, and therefore the initial work with information presented per API was continued by Region Stockholm. Originally, this information was presented in a printed brochure and on a webpage. In 2016, the information handled in Excel was transferred to a Structured Query Language (SQL) database and in 2019 to a Content Management System (CMS) database, the same one used by janusinfo.se. The database Pharmaceuticals and Environment gathers publicly available information about environmental hazard and risk. Some of the assessments have been generated by Region Stockholm or through academic experts. The work procedures and content of the database is based on a standard operating procedure for internal use. The document describes how environmental information is retrieved, processed, and presented; provides standard texts; suggests how to proceed if data are missing or incomplete; and suggests reference literature that can be used. Environmental information is presented per API. Environmental hazard includes data on persistence (P), bioaccumulation (B), and ecotoxicity (T) of the API. Briefly, each of the characteristics is assigned a numeric value, 0-3, depending on its ability to resist degradation in the aquatic environment, its ability to accumulate in adipose tissue of aquatic organisms, and its ability to poison aquatic organisms respectively [13,14,23]. The sum of the individual values, 0-9, makes up the hazard score. A higher hazard score indicates a greater potential to harm the environment. The T-value in the hazard score can refer to either acute or chronic ecotoxicity. We use a precautionary approach; thus, APIs with incomplete data are assigned a higher value. In the database, lack of data is indicated with an asterisk [13,14,23]. When publicly available, a conclusion on risk is presented. The risk is based on a comparison between the predicted environmental concentration (PEC) of an API and the concentration expected to be safe to aquatic animals and plants (PNEC, predicted no effect concentration) [13,[24][25][26]. PEC is calculated from sales figures in Sweden or predicted use in Europe. The risk can be stated as "insignificant," "low," "moderate," "high," or "cannot be excluded." "Cannot be excluded" is the term used when the pharmaceutical company has not provided enough data for a risk assessment. It is also stated if the API is exempted from risk assessment according to the guidelines of the Committee for Medicinal Products for Human Use (CHMP) [24]. The primary data sources for the database are the environmental information presented at fass.se and in the EPARs at EMA's website. The environmental information at fass.se is based on data from the pharmaceutical companies [27]. These data are reviewed by the consultancy firm the Swedish Environmental Research Institute [27]. There are EPARs for medicinal products approved from 2006 when the CHMP guideline was implemented. In general, we accept the hazard and risk assessments provided by EMA and LIF, but we are aware of the previous studies showing deficiencies in the data provided [28,29]. If environmental information is available both in EPARs and on fass.se, all is included in the database, and it is clearly stated on which data the overall assessment of hazard score and risk is based. The following aspects are considered when deciding this: the year and geographical area for the PEC-calculation, and the reliability and relevance of environmental fate and ecotoxicity studies. Additional data from peer-reviewed literature are used when assessed as reliable and relevant. Comparative environmental assessments for medically comparable alternatives can be performed by external experts. These assessments may differ from the assessments available at fass.se and EMAs website since they can be based on measured environmental concentrations (MEC) and include risk of selection of antibiotic-resistant bacteria [14,30]. When needed, academic experts are contacted for advice on how to interpret scientific studies and assessments, and the Swedish Medical Products Agency, EMA, and LIF are contacted for clarifications of available assessments. Assessing the use of the database The use of the database Pharmaceuticals and Environment was assessed using Google Analytics. Database visits were analyzed for the period January to September 2018, and the corresponding period in 2017. In October 2018, the technical platform for the database and janusinfo.se was changed which is the reason why the period thereafter was not included in the analysis. The following analyses were performed for the Swedish and the English version of the database: total number of page views, top three countries for traffic by countries, and top ten APIs. Identifying lessons learned from developing and managing the database During the years of developing and managing the database, different issues related to the function of the database as decision support have been discussed and handled. From these discussions, lessons learned have been phrased. The lessons concern presentation of information, access to data, handling discrepancies in data, comparison of data, and technical issues. This work has been done in collaboration between the person managing the database at Region Stockholm (H.R., pharmacist, PhD) and the database reference group Presentation of the database Pharmaceuticals and Environment The database currently (29 March 2019) contains information about 851 APIs. Of these, 154 APIs are exempt and 175 APIs lack information about P, B, and T. For 379 APIs, the risk is assessed as "cannot be excluded," i.e., data are missing or insufficient to calculate the risk. When updating the information in the database, we prioritize the APIs on the "Wise list" (the drug formulary of essential medicines for common diseases in Region Stockholm from the Drug and Therapeutics Committee). In addition, we also focus on other APIs widely used in Region Stockholm and APIs on Region Stockholm's list of 25 environmentally harmful APIs. This list was developed as a part of Region Stockholm's environmental program for the period 2017-2021, aiming to reduce negative environmental impact from APIs [31]. The expert panels of Region Stockholm Drug and Therapeutics Committee were involved in developing suggestions on how health care professionals can work to reduce environmental impact from the 25 APIs on the list. Their suggestions are included in the database. Figure 1 shows how environmental information is presented in the database, using the API paracetamol/acetaminophen as an example [14]. Firstly, information about the hazard score and risk is provided. This is followed by information about the T-value which is based on acute study data and the reference for the underlying data for P, B, and T. Below that, data from the references are presented in more detail (from fass.se for "Alvedon forte" in this case). There is also information presented about detected levels in Swedish water environment, including drinking water, and results from a comparative assessment of environmental risk for diclofenac, naproxen, ibuprofen, ketoprofen, etoricoxib, celecoxib, and Fig. 1 Information about paracetamol/acetaminophen from the database Pharmaceuticals and Environment at janusinfo.se [14]. P = persistence, B = bioaccumulation, and T = toxicity paracetamol/acetaminophen in Sweden. At the end, there is a reference list [14]. The API diclofenac provides an example of how extended information about an API in the database is presented [14]. Diclofenac is on Region Stockholm's list of 25 environmentally harmful APIs. The database contains information about diclofenac previously being monitored within the EU Water Framework Directive and may be proposed as a priority substance when the European Parliament and the Council propose a revised directive for priority substances in the EU Water Framework Directive. It also states that the Swedish Agency for Marine and Water Management has included diclofenac among special pollutants in its regulation. One can also read that diclofenac has been detected in treated wastewater and surface water in Region Stockholm. Results from a comparative assessment of environmental risk of diclofenac, naproxen, ibuprofen, ketoprofen, etoricoxib, celecoxib, and paracetamol/acetaminophen in Sweden are presented. For the health care professionals, there is a reminder that diclofenac is not recommended on the "Wise list." There is a concrete proposal that the health care professionals should review their recommendations to patients regarding overthe-counter analgesics. Finally there is a reference list [14]. Table 1 presents the total number of page views, page views for the top 10 APIs, and traffic by countries for 2018, for the Swedish and the English version of the database. Lessons learned from developing and managing the database Information can be missing, incomplete, or removed We have experienced that data provided in environmental risk assessments by EMA have been incomplete, e.g., lacked conclusion for bioaccumulation and/or persistence and the basis for calculating the environmental risk. When it is stated that CHMP had requested additional environmental information from a pharmaceutical company and the information could not be found at EMA's website, we have contacted the Swedish Medical Products Agency or EMA for clarifications. In some cases, the requested information had not yet been submitted, even after several years. In other cases, data had been submitted and were available upon request, or submitted but not available upon request. Furthermore, environmental information is presented per product at EMA's website, i.e., one must search through all approved medicinal products to find the available environmental information for a specific API. Fass.se is the main source of information for medicinal products approved before 2006. However, for a considerable number of these substances, there is no environmental information available. When available, ecotoxicity studies primarily concerned acute endpoints as chronic studies were not required prior to the implementation of the CHMP guideline. At fass.se, the environmental information for medicinal products needs to be updated every three years, otherwise it is removed. It is our experience that the information is sometimes removed instead of being updated. In those cases, we make use of previously published information that we have downloaded, and it is made clear in the database. A strategy for handling discrepancies in information is needed The environmental information provided by EMA and LIF may differ for different medicinal products with the same API, this problem has been pointed out before [29,32]. If there are different conclusions regarding an APIs ecotoxicity, persistence, or bioaccumulation, we present a worst-case scenario. If the PEC differs, we present a scenario with the most recent sales data for Sweden. Deviations from guidelines are commented on Deviations from the CHMP guideline and the guidance at fass.se have been found in the environmental risk assessments. For example, incorrect use of assessment factors for PNEC-calculations, or the use of ecotoxicity data with "greater than" (>) values. In those cases, the error is corrected, and/ or a comment is provided in the database. Different types of studies hamper comparisons of data When Region Stockholm initiated the database, ecotoxicity studies primarily concerned acute endpoints, but since the CHMP guideline was implemented, chronic ecotoxicity data have become available [24]. This is a welcome development, but it can hamper the comparison of the ecotoxicity between APIs. In the database, it is specified for each API if the value for T in the hazard score refers to acute or chronic studies. Following regulatory practice facilitates use It is desirable that the terminology in the database is consistent with the terminology used within the regulatory framework. The hazard score was previously called PBT index but was renamed in 2018 as PBT/vPvB (Persistent, Bioaccumulative, and Toxic/very Persistent and very Bioaccumulative) and is the term used within the regulatory framework for Substances of Very High Concern [24]. Further, the hazard score was initially developed to guide decision-makers not familiar with risk assessment procedures in the comparison between APIs. However, it has proven to be a rather blunt instrument. An API with a higher score is not necessarily of greater environmental concern. Therefore, there is an ongoing discussion within the reference group to remove the hazard score from the database. This would be in line with current regulatory practice. It is possible to generate and make use of new information If there is no risk assessment in the EPAR even though enough data are provided to make the calculation, the risk is calculated following the guidelines [24,33]. If ecotoxicity studies are missing at fass.se but have been published in previous environmental information at fass.se, the risk is calculated together with available sales data/PEC-value. When

deemed useful, an external expert has provided Region Stockholm with comparative environmental assessments on specific pharmaceutical groups, e.g., COX inhibitors, statins, and SSRIs. The purpose of these assessments is to clarify if any of the APIs within a pharmaceutical group is to be preferred from an environmental point of view. These investigations can be based on MEC and effect data for APIs, including risk of selection of antibiotic-resistant bacteria [14,30]. Available exposure data have been supplemented with data from Region Stockholm's screening program for APIs. It currently includes more than 100 APIs, measured primarily in the water environment [34]. In addition to the screening program carried out by Region Stockholm, other reports on pharmaceutical residues in water are also used as references for the database [35][36][37]. Transparency demands also apply to our work In the previous printed brochure, risk and hazard scores including numeric values for the individual characteristics of P, B, and T were presented without the underlying data. To increase the transparency, underlying data for hazard assessment are now successively added to the information for individual APIs, including references for the information, see Fig. 1. Furthermore, the underlying data for the exposure assessment, and consequently the conclusion on risk, are included. This is important since the exposure assessment for different APIs may be based on PEC (using either expected use or sales data from different years), or on MEC. The increased transparency facilitates critical evaluation of the assessment and comparison between APIs. A user-friendly technical platform is a crucial key The transfer of the environmental information to a database using a Content Management System offers benefits to users. More comprehensive information can be presented, and a reference list is included. The content is kept more up to date as compared with the printed folder. For the editor, the transfer from the Excel-based system facilitates handling of more information. Discussion Decision support on environmental effects for pharmaceuticals is needed, both in the private and public sector, to address the potential problems related to production and use of pharmaceuticals. It is our opinion that decision support should primarily be provided by regulatory agencies, on national (e.g., the Swedish Medical Products Agency) or EU-level (e.g., the EMA or the European Commission). There are two reasons for this: their presumed objectivity and their access to public and confidential information. Currently, the Swedish Medical Products Agency does not provide any decision support and, as shown in this paper, the information provided by EMA is not always sufficient to serve as decision support. The development of other databases also indicates that the decision support provided has not been sufficient. It is our interpretation that this absence of adequate decision support relates to that environmental effects from pharmaceuticals is not a field of particular interest to decision-makers. This can be seen both when comparing the management of pharmaceuticals with other chemicals and comparing "environmental side effects" with medical side effects of pharmaceuticals. We think this conclusion is supported by several observations. The first legal requirement for environmental risk assessment of pharmaceuticals was established relatively late [38] and the requirements are low in comparison with those for other chemical groups. This is illustrated by the fact that the requirements only apply to medicinal products seeking marketing authorization after the legislation was brought into force. Also, the legal demands on companies to provide the requested environmental information appear to lack power as environmental information is still missing several years later. Other important processes related to environmental effects from pharmaceuticals have also been delayed or ignored. The EU's "Strategic approach to pharmaceuticals in the environment" was delayed by several years, and no actions have been taken to deal with the severe emissions from production of pharmaceuticals for the EU market [39]. Another indication of the indifference is that environmental effects are not included in the monitoring of the safety of pharmaceuticals that is carried out by manufacturers and regulatory agencies throughout their life cycle. EU law requires marketing authorization holders, national competent authorities, and EMA to operate a pharmacovigilance system, but that does not include environmental aspects. Just as there can be new studies published on medical effects and side effects of pharmaceuticals after approval, new environmental studies could alter the assessments. The database Pharmaceuticals and Environment at janusinfo.se presents compiled, comprehensive, and easily accessible environmental information about APIs. This information can be used when communicating about environmental risks, and as decision support when assessing whether it would be favorable, from an environmental point of view, to substitute one API for another. The content in the database has a Stockholm and Sweden perspective in terms of use (sales data) and MEC but hazard assessments are primarily based on non-site-specific tests. Other medical knowledge databases within Region Stockholm, e.g., the databases Drugs and Birth Defects, and Interactions [40,41], are designed to be decision supports in every-day clinical work. In contrast, the database Pharmaceuticals and Environment offers the Stockholm Drug and Therapeutics Committee, and other Swedish Drug and Therapeutics Committees, a possibility to include environmental aspects when recommending pharmaceuticals for health care providers [42]. Many Drug and Therapeutics Committees in Sweden include environmental aspects when making pharmacotherapeutic recommendations. These recommendations help physicians and other prescribers to make good environmental choices without compromising medical effectiveness and safety when prescribing pharmaceuticals [42]. Since the database contains suggestions on how to reduce the environmental impact from particularly environmentally harmful APIs, the database can also be used as a management tool by health care providers. Also, researchers and wastewater treatment operators can make use of the compiled environmental information. To increase our understanding of the current use and how the database could be further developed, it would be interesting to perform a survey among users. From the lessons learned, we conclude that one major problem is lack of environmental information for a majority of APIs, not least among the medicinal products approved before 2006 [24]. This has also been noted by the European Commission and others [19,39,43]. For medicinal products approved before 2006, the database is dependent on the information available through fass.se and the peer-reviewed literature. A drawback with older information is that it mainly reflects the acute ecotoxicity of the APIs. In addition, since it is optional for companies to publish information at fass.se, medicinal products may lack environmental information completely. Also, environmental information that has been published can be withdrawn, which makes fass.se a precarious source of information. It has been estimated that 5-10% of medicines in use might pose an appreciable risk to the environment [32,44]. A pragmatic way to identify these substances for further environmental assessments could be to apply a prioritization approach [44]. From an environmental point of view, it is desirable to find medically comparable alternatives that are less environmentally harmful. In decision-making, it is not enough to know that one API has a negative environmental impact if data are missing for the alternatives. The assessment of whether substituting one API for another is environmentally beneficial or harmful also poses a challenge. For the environmental risk assessments presented by EMA and LIF, it would be desirable that the complete environmental risk assessment, including underlying studies, is made publicly available and easily accessible. At present, there is no regulation regarding which information that needs to be made public in the EPARs. CHMP's interpretation of data could also be more perspicuous; there are for example several EPARs communicating test results for persistence without a conclusion. It is desirable that CHMP develops a process to ensure that all requested information in the environmental risk assessments is submitted by the manufacturers. Furthermore, it would be desirable that environmental data are presented in a compiled assessment per API and that all the information is easily accessible at EMA's website. A change in consumption of a medicinal product can alter the calculated risk, and new knowledge could affect the conclusions in a previously performed environmental risk assessment. It would therefore be valuable if the environmental information for APIs is updated at regular intervals, which is standard procedure in regulatory frameworks for other chemicals [38]. However, since the benefit/risk assessment for human medicinal products at present does not include environmental effects, an update of the environmental risk assessment is not required for renewals of marketing authorizations [24]. We welcome the proposal for a revised EMA guideline [45]. The proposal includes a tailored risk assessment for APIs with specific classifications (e.g., endocrine active substances, antibiotic substances) and further tests, e.g., the estimation of the exposure of predators to pharmaceuticals through the food chain ("secondary poisoning"), as well as directly through the environment. If the guideline were to be updated with the proposed changes, it would improve the information in the environmental risk assessments, and hence in the database. In the absence of an adequate decision support produced by the regulatory authorities, we find the database Pharmaceuticals and Environment to be a useful decision support for Swedish Drug and Therapeutics Committees and health care providers, and it is our belief that the information can be valuable also in other settings. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The impact of maternal health care utilisation on routine immunisation coverage of children in Nigeria: a cross-sectional study Objective To examine the impact of maternal healthcare (MHC) utilisation on routine immunisation coverage of children in Nigeria. Design Individual level cross-sectional study using bivariate and multivariable logistic regression analyses to examine the association between MHC utilisation and routine immunisation coverage of children. Setting Nigeria Demographic and Health Survey 2013. Participants 5506 women aged 15–49 years with children aged 12–23 months born in the 5 years preceding the survey. Primary outcome measures Fully immunised children and not fully immunised children. Results The percentage of children fully immunised with basic routine childhood vaccines by the age of 12 months was 25.8%. Antenatal care (ANC) attendance irrespective of the number of visits (adjusted OR (AOR)1–3 visits 2.4, 95% CI 1.79 to 3.27; AOR4–7 visits 3.2, 95% CI 2.52 to 4.13; AOR≥ 8 visits 3.5, 95% CI 2.64 to 4.50), skilled birth attendance (SBA) (AOR 1.9, 95% CI 1.65 to 2.35); and maternal postnatal care (PNC) (AOR 1.7, 95% CI 1.46 to 2.06) had positive effects on the child being fully immunised after adjusting for covariates (except for each other, ie, ANC, SBA and PNC). Further analyses (adjusting stepwise for each MHC service) showed a mediation effect that led to the effect of PNC not being significant. Conclusions The percentage of fully immunised children in Nigeria was very low. ANC attendance, SBA and maternal PNC attendance had positive impact on the child being fully immunised. The findings suggest that strategies aimed at maximising MHC utilisation in Nigeria could be effective in achieving the national coverage target of at least 80% for routine immunisation of children. Many times throughout the manuscript, authors use either reveal or revelation! Can they elaborate about what really meant in using insistently these words because all findings reported here are well known? Specific comments: • However, I couldn't see any efforts to identify mechanisms through routine immunization can really affect children's health. Is routine immunization a panacea? Or are there any other factors (demographic, social, economic, cultural) which can interfere with routine immunization to be beneficial to children's health? • On another note, results are presented without any efforts to explain what is happening and how these findings can really help to advance our knowledge on this matter and policymakers. They present grossly findings without any effort of interpretation in Nigerian context. • Methodologically, I have some reserves regarding the estimates presented in the manuscript: o

Am I right if I say that estimates presented here are based on a sub-sample of women with children 12-23 months born in the five years preceding the survey? o Relatedly, the authors recognize that NDHS used a complex survey design (CSD), but then didn't elaborate on how CSD may affect estimates, especially confidence intervals (CI) o From the 2 points above, I am unable to judge the quality of the estimates presented here. Indeed, because estimates are based on a sub-sample, authors should clearly state how they took into account the (i) CSD of NDHS; (ii) the fact that estimates are based on a sub-sample. o Section on study design and data collection is unnecessary because the study is based on secondary data analysis • Results o Table 1-Because the dependent variable is binary, it is unnecessary to present both not fullyand -fully immunized. 2. Abstract L5 page1 (P) Objective does meet the main findings and conclusions. L9 Design -vague. That is, mixed with analysis L22 Results-The percentages of children fully immunized is unclear. Why only for 12 months age? What about percentages for those up to the ages of 23 months? Or Does it mean percentage of children fully immunized with basic child hood vaccines by the age of 12-23 months was 25%? Please, clarify it L3 p-Conclusions…and maternal PNC attendance had positive impacts on the child being fully immunized……does not seem results of a cross-sectional design. 3. Strength and limitations page 2 -Needs revision. Sentences should be kept to the reasonable length in this section as per the journal requirements L53 page 2-the reliability of this study is increased by the suitability of research questions, the quality of the NDHS women's questionnaire, having well categorized dependent variable, etc are all not clear. And also this is not a bulletin rather it is a paragraph. What is well categorized dependent variable? L17 page 3-The NDHS questionnaires are not self-administered--which makes the data prone to interviewer administered ….is remains unclear…Is that a self-administered questionnaire that is prone to interviewer bias? In general, this part needs to be summarized 4. Introduction pages 4-5 L14 …….mortality to about 25/1000live births by 2030. [1, 2, 3] seems wrong full stop. I suggest "…..mortality to about 25/1000live births by 2030 [1,2,3]. "is correct as per the journal style. Same for all references in the document. 5. Factors associated page6-7 This heading is extra and gives no relevant information since it can be summarized under introduction heading L12 studies have shown…….healthcare of the newborn. (lacks citation) L16-to be fully immunized in Ghana, the Gambia, Burkina Faso,….citations required close to each country ( not often suggested to merge them) L49-51-authors should provide strong justifications on the significance of this study in Nigerian context, since prior studies have been conducted on this issue, using each separate components of MHC service, according to this paragraph description. 6. Methods L19-25 p7-Nigeria is a sub-Saharan Africa----->with 774 local government areas narrates something about a study area (Nigeria) L29-43 p7 To examine the impact…..>(2638, 2868 males) were included in this study provides descriptions of a study population. The question is what was the study setting? Is it a communitybased? Institution/facility based or what? Who are the target populations? From how many reproductive age women and how did you calculate/draw a sample of 5506 women for inclusion? What is your decision point to exclude the remaining reproductiveage women in the area? This is the major limitation L12 p8-How did you draw a fixed sample unit of 45 households? L41 p9 Other form of birth attendance....the term" other form" should be described/lists are recommended and also similar concern for L10 p10 In general, the discussion should be concise and kept to a reasonable length 9. Strength and limitations lL43 p21.....proper formation of sub-categories of the predictor variable..provides no statistical evidence. 10. Conclusions L3-8 p23...the conclusion ANC attendance of the mother, mothers' use of SBA,......had positive impact on the child being fully immunized.... is better described as a finding from cross-section. That is, ANC attendance of the mother, mothers' use of SBA,......were significantly associated with the child's full routine immunization. L8-12 p23....better to improve the recommendations with respect to each predictor variable examined to have had significant association and should able to provide a clue for public health practices 11. References-provide web links for some references e.g . Ref. 3,7,8,21,22,29,31,34,& 44 REVIEWER Minjin Kim University of Massachusetts Medical School, United States REVIEW RETURNED 25-Jan-2019 GENERAL COMMENTS Thank you for inviting to review this manuscript. I have reviewed it carefully with a great deal of interest as the emerging literature shows that immunization is an efficient and cost-effective public and global health intervention. This manuscript explored the impact of maternal health care utilization (MHC) on routine immunization coverage of children Nigeria by exploring 1) the association between routine immunization coverage and antenatal care attendance, skilled birth attendance, and maternal postnatal care and 2) the relationship between routine immunization coverage, predictor variables, and covariates. This manuscript is well written, and the results are very interesting and convincing. The strength of the manuscript is that this is the study results are based on the Nigeria Demographic and Health Survey, which is a nationally representative survey that provides a large sample size. The limitation of the study is that the children's immunization status was from verbal reports of the mothers, which is addressed in the text. While the manuscript is very interesting and clear in delivering the message of the importance in maximizing MHC in Nigeria, I would like to suggest including the potential strategies or programs in relation to policy, education, and research that might aid in maximizing MHC in Nigeria. In addition, I would like to encourage authors to consider including a theory to frame the research questions and explain the relationship. Again, I appreciate the opportunity to have read your work. I look forward to seeing more of it in the future. All the best and beyond! Catherine Ford University of Illinois -Chicago, United States REVIEW RETURNED 31-Jan-2019 GENERAL COMMENTS Overall: This article is about the association between maternal health care utilization and uptake of childhood immunizations. The article is well written on important topic. There are a handful of grammatical errors that should be addressed. -Why is there a section of strengths/limitations between the abstract and the manuscriptthis is already included in the discussion and repetitive Introduction: -The second paragraph could be more succinct. You discuss the history of immunizations until 1990 in several sentences, which could be collapsed addressing the main point. And then there is no mention of anything between 1990 and 2013 except for that there was minimal improvement. If anything, this timeframe could be a bit more developed than the historical part. -You repeat the exact immunizations in paragraph two and three, this is redundant, you can state (i.e. children who received all six recommended vaccines) and leave it at that. -You don't need to have the sub-header "Factors associated with immunization coverage and significance of the study". This is implied by it being the last paragraph of the introduction. Methods -In the study design sub-section, you should include the permission of using the NDHS as well as the ethical approvals so that this is included in the body of the manuscript -In the dependent variable sub-section, the definition of "not fully immunized" is confusing. It reads as if this group includes only children who did not get any immunizations or only got immunizations after 12 months old. What if a child got one in the series, or got all of one vaccine but not any of the others. Can you please clarify this. -Can you clarify inclusion and exclusion criteria because that might better clarify the "fully immunized" to the "not fully immunized" participants -Can you clarify if there were any women who may have been included twice (a 12 month old and a 23 month old) and how that was avoided -You don't need to include how the variables were coded (the last sentence of the dependent variable sub-section) -In the sub-section Statistical Analysis/Bivariate Analysis, why did you choose p<0.25 to include in the multivariable logistic regression? -In the sub-section Statistical Analysis/Multivariable Logistic Regression Analysis, there is a significant amount of detail in everything that was done, but this is included in Figure 3 and 4. You may want to be more succinct and either have an appendix that explains this further or direct readers to the figures to explain this further. -The sub-section Patient Involvement can be included in the Study Design subsection in one sentence, it does not need to be it's one sub-section Results -Overall very clear, may consider being more succinct as most of what is included is clearly displayed in the figures. For example the first part of the Multivariable logistic regression results subsection repeats the same variables that were adjusted for three times, it could be all stated once and then refer back to these variables. Discussion -In paragraph 2, you state there is a differential between countries regarding the impact of ANC on routine immunizations, but you do not give any actual numbers to make that statementcan you do so? -Paragraph talks about the importance of the 8 visit model for ANC and how that can improve the immunizations, however, you don't discuss the fact that 40% of respondents got no ANChow will that population go from 0 to 8? -In limitations, it would be important to add to recall bias the possibility that there could be confusion between last pregnancy and the pregnancy before the last one if this is including data from 5 years prior to 2013. It is likely individuals could be pregnant multiple times in that time frame Comment 5: Page 7, row 40: It is advisable to author to include the brief explaination or sample size calculation on how the 5506 women was an ideal size. Response: The survey is a nationwide survey and the sample size is much greater than the minimum number required for the study, therefore no justification is needed in this regard. Results: Comments 6: Page 14, row 31: Highlighted to main findings of the sex: such 74.4% and 73.9% will do as the table 1 already self explanatory. Response: We understand the reviewers perspective that Table 1 is self-explanatory regarding the descriptive analysis results. However, we find it is still necessary to provide the actual numbers together with the percentages to help the reader by maintaining the flow of the text because the percentages alone in the main manuscript do not provide much information. Discussion: Comment 1: The paper is well written and concise. I commend that the authors did a good job to explain the link between routine immunization and vaccine-preventable diseases which ultimately reduce neonatal and under-five mortality in developing countries and in Nigeria in particular. Many times throughout the manuscript, authors use either reveal or revelation! Can they elaborate about what really meant in using insistently these words because all findings reported here are well known? Response: The word "revealed" is used to denote what the study found. However, it has been replaced with the word "showed" throughout the text to increase understanding of the text. Comment 2: However, I couldn't see any efforts to identify mechanisms through routine immunization can really affect children's health. Is routine immunization a panacea? Or are there any other factors (demographic, social, economic, cultural) which can interfere with routine immunization to be beneficial to children's health? Response: Other factors that can affect routine immunization have been mentioned under "Introduction" in the last paragraph (See page 5: lines 18-23; and page 6: lines 1-3). Comment 3: On another note, results are presented without any efforts to explain what is happening and how these findings can really help to advance our knowledge on this matter and policymakers. They present grossly findings without any effort of interpretation in Nigerian context. Comment 5: Relatedly, the authors recognize that NDHS used a complex survey design (CSD), but then didn't elaborate on how CSD may affect estimates, especially confidence intervals (CI) Response From the 2 points above, I am unable to judge the quality of the estimates presented here. Indeed, because estimates are based on a sub-sample, authors should clearly state how they took into

account the (i) CSD of NDHS; (ii) the fact that estimates are based on a sub-sample. Response: Measures that were taken to prevent underestimation of the standard errors as a result of the complex survey design have been provided under "Statistical analysis", sub-section "Preliminary data analysis" (See page 11: lines 7-10). Comment 6: Section on study design and data collection is unnecessary because the study is based on secondary data analysis Response: Though we use secondary data, we still deem it necessary for the readers to have complete understanding of the methods used in this study. That is why the section on study design is provided accordingly. Also, some readers may be familiar with DHS dataset and approach but not all fall into this category. Thus on the contrary to the point raised, we still find it necessary to include the study design irrespective the kind of data used in the study. Additionally, the study design does not only talk about the NDHS, but also about the design of this study. Results: Comment 7: Table 1-Because the dependent variable is binary, it is unnecessary to present both not fullyand -fully immunized. Response: We understand the point of view of the reviewer. However we think that including both groups makes it easier for readers to grasp the information presented in the crosstabulation with less difficulty, though we agree that same information can be calculated giving the total numbers and the values in one arm. REVIEWER 3 Reviewer Name: Tesfaye Hambisa Mekonnen Institution and Country: University of Gondar, College of Medicine and Health Sciences, Ethiopia Comment 1: The authors tried to address the impact of MHC service usage on routine immunization of children in Nigeria using a cross-sectional study. Such study is relevant because uptake of immunization among children in majority of the developing countries, such as Nigeria, yet remains lower than the expected national coverage target of at least 80% in accordance with WHO initiatives. But I am afraid of the manuscript is suitable for publication in its current form. Comment 2: Several major issues need to be addressed and thoroughly revised before considering it for publication. -The authors should follow the journal's manuscript report formats and styles For example, they should include "Main text" heading next to introduction before the heading "Methods". -Keywords are not placed appropriate. They should come after abstract next to conclusion, before strength and limitation heading -The manuscript document is not well structured/organized. Response: The manuscript has been formatted according to the journal's style and all irregularities have been checked and corrected accordingly. Comment 3: I don't think the authors are expected to report each and every thing done, rather they should give a concise description based on their objectives For example, the heading"Factors associated with immunization coverage and significance of the study" could be summarized under introduction section, the subheadings dependent variables, predictor variables….and controls can be summarized under a single heading (for example, you can say 'variable measurements', all the subheadings under statistical analysis and before patient involvement section can also be organized under a single heading for example, you can say, 'Data /or statistical analysis ', result section can be summarized based on the main objectives of the study as descriptive and analytical (factors associated) sections each separately. This is important to be done since it keeps the manuscript to a reasonable length Requested suggestions/detail Response: The sub-header "Factors associated with immunization coverage of children and significance of the study" has been removed. The section has been converted to paragraphs 4 and 5 under "Introduction". Additionally, these paragraphs (4 and 5) have been summarised to avoid unnecessary repetition. However, individual sub-headings for the dependent and predictor variables have been maintained to enhance clarity. Comment 4: The title requires little modifications. In its current form, it does not exactly reflect what has been presented by the paper. For example, the title tells about the impact, whereas the finding and analysis conclude about associations between MHC usage and level of immunization uptake. -Do the authors think impact and associations are similar? -Can investigating immunization level and the factors influencing it simultaneously using a cross-section design help concludes the impacts of one variable on another? If so, what a specific type of analysis was employed to evaluate impact than the analysis used in this study to investigate associations? If the objective is mainly to evaluate the impact, the method employed in the current study does not seem appropriate since impact evaluation requires some intervention/observation/longitudinal /prospective follow up study. Age of the children also needs to be mentioned in the title. I suggest the title could be: Associations of maternal healthcare utilization and routine immunization coverage among children aged 12-23 months in Nigeria: A cross-sectional study Response: We agree with the reviewer that cross-sectional studies are generally limited in its ability to infer causality from an association mainly because of the issue with temporality between the exposure and outcome. However, that is not so for all kinds of exposure/outcome relationships. To the best of the authors' knowledge, analytical cross-sectional studies can be used to infer impact in this case because the exposure (maternal health care utilization) happens before the outcome (routine immunization). In other words, no child can be immunized (in his mother's womb) before maternal health care attendance of the mother. That makes it plausible in our study to make such inferences though we agree that one generally has to do so with caution. Also, in another sense of the word "impact" denotes effect, not causation, hence, it is used even in titles for qualitative studies. Furthermore, we don't see any added benefit of add much details in the title. Routine immunization is known for children under 1 year old. Children aged 12-23 months is not the age bracket for routine immunisation. It is just an age bracket of children selected for this study. Therefore, we deem it not necessary to add the age bracket of children selected for the study in the title. The same goes with the age bracket of mothers selected for this study. Abstract: Comment 5: L5 page1 (P) Objective does meet the main findings and conclusions. L9 Design -vague. That is, mixed with analysis L22 Results-The percentages of children fully immunized is unclear. Why only for 12 months age? What about percentages for those up to the ages of 23 months? Or Does it mean percentage of children fully immunized with basic child hood vaccines by the age of 12-23 months was 25%? Please, clarify it L3 p-Conclusions…and maternal PNC attendance had positive impacts on the child being fully immunized……does not seem results of a cross-sectional design. Response: Though we very much value the comments provided, we also notice that there is a misunderstanding regarding the exposure and outcome measures that we used in this study and the naturally occurring temporal gap (temporality) between them. The comment regarding the use of the word impact and making such inference has been addressed in the response above. Children aged 12-23 months is not the age bracket for routine immunisation. It is just an age bracket of children selected for this study. Strength and limitations page 2: Comment 6: Needs revision. Sentences should be kept to the reasonable length in this section as per the journal requirements L53 page 2-the reliability of this study is increased by the suitability of research questions, the quality of the NDHS women's questionnaire, having well categorized dependent variable, etc are all not clear. And also this is not a bulletin rather it is a paragraph. What is well categorized dependent variable? L17 page 3-The NDHS questionnaires are not self-administered---which makes the data prone to interviewer administered ….is remains unclear…Is that a self-administered questionnaire that is prone to interviewer bias? In general, this part needs to be summarized Response: According to the journal´s instruction in citing in the text, "reference numbers in the text should be inserted immediately after punctuation, with no word spacing". Factors associated page 6-7: Comment 8: This heading is extra and gives no relevant information since it can be summarized under introduction heading L12 studies have shown…….healthcare of the newborn. (lacks citation) L16-to be fully immunized in Ghana, the Gambia, Burkina Faso,….citations required close to each country ( not often suggested to merge them) L49-51-authors should provide strong justifications on the significance of this study in Nigerian context, since prior studies have been conducted on this issue, using each separate components of MHC service, according to this paragraph description. Response: The sub-header "Factors associated with immunization coverage of children and significance of the study" has been removed. The section has been converted to paragraphs 4 and 5 under "Introduction". Additionally, these paragraphs (4 and 5) have been summarised to avoid unnecessary repetition. Methods: Comment 9: L19-25 p7-Nigeria is a sub-Saharan Africa----->with 774 local government areas narrates something about a study area (Nigeria)L29-43 p7 To examine the impact…..>(2638, 2868 males) were included in this study provides descriptions of a study population. The question is what was the study setting? Is it a community based? Institution/facility based or what? Who are the target populations? From how many reproductive age women and how did you calculate/draw a sample of 5506 women for inclusion? What is your decision point to exclude the remaining reproductive age women in the area? This is the major limitation L12 p8-How did you draw a fixed sample unit of 45 households? Response: A secondary data (NDHS 2013) was used in this study. A summary of how 45 households were selected has been given under sub-section "Study design and data collection". More details regarding the sampling design can be found in the NDHS 2013 report. Comment 10: L41 p9 Other form of birth attendance....the term" other form" should be described/lists are recommended and also similar concern for L10 p10 Response: e reason why p<0.25 was chosen was already provided in paragraph 2 under "Statistical analysis" sub-section "Bivariate analysis". However, it has now been rephrased for easy understanding (See page 12: lines 5-8). Clarification has been provided regarding the stepwise regression (See page 12: lines 12-13). The amount of information about multivariate analyses under "Statistical analysis" sub-section "Multivariable logistic regression analysis" has been reduced (See page 12: lines 20-23; and page 13: lines 1-2), because the information about the analysis is already provided in Table 3 and Response: The numbers and percentages under "Not Fully Immunized and Fully Immunized" were calculated while applying the weight whereas this is not the case for calculations under "Total Population". The weight is applied in all analysis involving significance testing and confidence intervals to restore the representativeness of the sample (See page 11: lines 7-9). Therefore, the percentages of not fully immunized and fully immunized male children were calculated with respect to the total weighted number of male children (See table 1 Response: The study was done using data from in many low-and lower-middle income countries. Mentioning all the countries from where the data was collected could lead to so much details in the manuscript. More details about the study can be found in the articles. Comment 16: In general, the discussion should be concise and kept to a reasonable length Response: Making the discussion more concise than the way it is will lead to superficial discussion which will lead to the loss of important issues that should be discussed. Comment 1: Thank you for inviting to review this manuscript. I have reviewed it carefully with a great deal of interest as the emerging literature shows that immunization is an efficient and cost-effective public and global health intervention. This manuscript explored the impact of maternal health care utilization (MHC) on routine immunization coverage of children Nigeria by exploring 1) the association between routine immunization coverage and antenatal care attendance, skilled birth attendance, and maternal postnatal care and 2) the relationship between routine immunization coverage, predictor variables, and covariates. Comment 2: This manuscript is well written, and the results are very interesting and convincing. The strength of the manuscript is that this is the study results are based on the Nigeria Demographic and Health Survey, which is a nationally representative survey that provides a large sample size. The limitation of the study is that the children's immunization status was from verbal reports of the mothers, which is addressed in

the text. While the manuscript is very interesting and clear in delivering the message of the importance in maximizing MHC in Nigeria, I would like to suggest including the potential strategies or programs in relation to policy, education, and research that might aid in maximizing MHC in Nigeria. In addition, I would like to encourage authors to consider including a theory to frame the research questions and explain the relationship. Response: A paragraph summarising ways to maximize MHC utilization in Nigeria has been added under conclusions. Regarding the inclusion of theory to frame the research question, the theoretical perspective is already implied in the introduction when it is stated that routine immunization is integrated into maternal health care because of its potential to increase routine immunization coverage in the country. Hence, the research question is framed based on this conceptual understanding. Comment 3: Again, I appreciate the opportunity to have read your work. I look forward to seeing more of it in the future. All the best and beyond! REVIEWER 5 Reviewer Name: Catherine Ford Institution and Country: University of Illinois -Chicago, United States Overall: Comment 1: This article is about the association between maternal health care utilization and uptake of childhood immunizations. The article is well written on important topic. There are a handful of grammatical errors that should be addressed. Response: Grammatical errors in the text have been checked and corrections made accordingly. Comment 2: Why is there a section of strengths/limitations between the abstract and the manuscriptthis is already included in the discussion and repetitive Response: The section on strengths and limitations provided between the abstract and the manuscript has been done in accordance with the journal requirements. Introduction: Comment 3: The second paragraph could be more succinct. You discuss the history of immunizations until 1990 in several sentences, which could be collapsed addressing the main point. And then there is no mention of anything between 1990 and 2013 except for that there was minimal improvement. If anything, this timeframe could be a bit more developed than the historical part. Response: The second paragraph has been collapsed a bit. Also, effort has been made to retain the main information in the second paragraph. Additional information about reduction in neonatal mortality rate from 1990-2013 has been added in the second paragraph (See page 4: lines 20-23; and page 5: lines 1-3). Protected or Unprotected Sex: The Conceptions and Attitudes of the Youth in Bolgatanga Municipality, Ghana The youth in Bolgatanga municipality in Ghana have relatively less knowledge of sexual and reproductive health (SRH) compared to the youth in other parts of Ghana. More fundamental knowledge is needed of the factors that influence young people to have protected and unprotected sex in specific social and cultural contexts, in order to protect them from adverse consequences, such as sexually transmitted diseases (STIs), HIV/AIDS and unintended pregnancies. This study therefore analyzed the conceptions and attitudes of the youth toward protected and unprotected sex, and particularly condom use, in Bolgatanga municipality. Semi-structured and focus group interviews were held with 71 young males and females and 17 adults. The results indicated that many of them lack a comprehensive knowledge of STIs, contraceptives and pregnancy, while a group of them had a negative attitude towards contraceptives. Not all parents, schools and organisations provide young people with a comprehensive education about SRH, and some even discourage such education because they believe it would encourage young people to have sex before marriage. In addition, young people also inform each other about SRH issues, sharing stories and personal experiences with their peers. The information they exchange is not always correct, however; sometimes it merely reflects their own personal preferences. The unequal power in the sexual relationships of young people—related to the traditional value system that gives men, but not women, “sexual freedom, both in and outside marriage”—is another determining factor for unprotected sex. Introduction . However, a substantial proportion of them lack sufficient knowledge about the use of contraceptives, and use contraceptives inconsistently (Ohene and Akoto 2008). Condom use among sexually active young people is low in Ghana (Ghana Aids Commission 2014; Krugu et al. 2016b). For example, in 2014, only 32% of sexually active unmarried females aged 15-19 years used modern contraceptives (see Table 1; Ghana Statistical Service, Ghana Health Service and ICF International 2015).Condom use was reported by less than 40% of males in 2008 (Doyle et al. 2012). There are various reasons for this, such as their low risk perception, a lack of adequate access, and their concerns about costs and confidentiality (Awusabo-Asare et al. 2006;National Population Council 2011;Bankole et al. 2007;Awusabo-Asare et al. 2004). Additionally, condom use is seen as unnatural and unpleasant, and it may cause distrust in sexual relations (Kuumuori Ganle et al. 2012). It was reported that the higher a woman's education, the more likely she is to ask her partner to use condoms. Rich or middle income women are more likely to ask their partners to use condoms compared to poorer women (Darteh et al. 2014). It is unclear whether transactional sex (i.e., sex for money) affects condom use (Moore et al. 2007). Most churches and religious groups do not promote condoms, because they fear that condom use promotes promiscuity (Appiah-Agyekum and Suapim 2013). It should be remarked that it is not clear if the mentioned studies concerning condom use focus on transmission of HIV/AIDS and other STIs through vaginal intercourse only, or include oral and anal sex as well. Several scholars have argued that more fundamental knowledge is needed about factors that influence the unprotected and protected sexual behavior of young people, and about the role of these hindering and motivational factors in various social and cultural contexts. This insight could contribute to the development of more tailored and effective sexual and reproductive health (SRH) programs to protect young males and females from the adverse consequences of having sex without contraceptive methods (Darteh et al. 2014;Karim et al. 2003;Awusabo-Asare and Annim 2008;Madise et al. 2007). The youth in Bolgatanga municipality, the capital of the Upper East Region, have relatively less knowledge of SRH, including STIs and HIV/AIDS, compared to the youth in other parts of Ghana, and they have a relatively low level of familiarity with family planning methods ( Van der Geugten et al. 2015;Rondini and Kingsley Krugu 2009). A recent study in Bolgatanga municipality reported that 24% of sexual active junior high school students had used a condom the last time they had sex (Krugu et al. 2016b). However, compared to other parts of Ghana, research in this rural, relatively remote northern area is limited. Abstinence from premarital sex is prescribed for the Frafra (in particular the Gurune) and the Christian and Islamic groups in Bolgatanga municipality, and the virginity of unmarried women is highly valued (Van der Geugten et al. 2013). A considerable number of young males and young females are, however, sexually active before marriage, influenced by increasing modernization, education, and new media (Ghana Statistical Service, Ghana Health Service and ICF Macro 2009; Krugu et al. 2016b; Van der Geugten et al. 2013). In the Upper East Region, the median age of marriage in the region is 24.4 years for males and 18.9 years for females. For females, there is only a slight difference between the median age of first sexual intercourse (18.4) and the median age of marriage (18.9). For males the difference is larger: their median age of first sexual intercourse is 21 and their median age of marriage is 24.4 (Ghana Statistical Service, Ghana Health Service and ICF International 2015). Quantitative research concerning the sexual behaviour of youth and contraceptive use was done in various parts of Ghana, and recently in Bolgatanga municipality as well (Krugu et al. 2016b;Ghana Statistical Service, Ghana Health Service and ICF International 2015;Ohene and Akoto 2008). However, qualitative research is recommended to gain more fundamental knowledge in specific social and cultural contexts with respect to contraceptive use, STIs (including HIV/AIDS), and unintended pregnancies. Therefore, in the present study, a qualitative method was applied to (1) investigate the conceptions and attitude of the youth toward unprotected and protected premarital sex, in particular condom use, and (2) investigate the sociocultural dynamics and context of using or not using contraceptives as a youth, and the risk of the youth for HIV/AIDS and STIs. The aim is to gain a better understanding of what drives the youth in Bolgatanga municipality in northern Ghana to practice protected or unprotected sex. Method Design Semi-structured interviews were carried out with 32 young males and females, and focus group interviews were conducted with a further 39 young males and females. Individual interviews provided a safe setting and privacy for the respondents, while the focus groups motivated respondents to share their ideas and to react to each other. To answer the second research question, not only interviews were held with young males and females, but also with 17 adults who were familiar with the local youth and their problems. These respondents (mostly teachers, social workers, and health workers) provided complementary information about the sociocultural dynamics and context of using or not using contraceptives as a youth, and the risk of the youth for HIV/AIDS and STIs. Setting Ghana has almost 25 million inhabitants, divided over 10 regions. The three northern regions are the poorest. They are mainly rural and the majority of the people live in villages or small communities. The main source of income is farming. School attendance and literacy rates are low compared to the rest of Ghana (Ghana Statistical Service 2012). Bolgatanga municipality (132,000 inhabitants) is the capital of one of the northern regions. The dominant ethnic group in Bolgatanga municipality is the Gurune (Ghana Statistical Service 2005). The three main religions in Bolgatanga municipality are traditionalism (practised by 22.3% of the population), Christianity (57.6%) and Islam (17.1%). Only 2.7% have no religious affiliation (Ghana Statistical Service 2014). The availability of contraceptives in Bolgatanga municipality varies per area. Various brands of male condoms are available at pharmacies, supermarkets, and health clinics, mainly in Bolgatanga town, but also in some of the rural communities. The average price of a condom is US$ 0.07 (Ghana Statistical Service, Ghana Health Service and ICF Macro 2009). Female condoms were not available in Bolgatanga municipality during data collection. Contraceptive pills and contraceptive injections are available free of charge at government hospitals and clinics upon presentation of a doctor's prescription. Contraceptive pills are also sold at pharmacies (the average price of a month's worth of pills is US$ 0.09). Population and Sampling The research population comprised young males and females (14-25 years, varying levels of education) and adults (various ages, various background, occupations and religions). For both young people and adults, snowball sampling was done, taking into account gender, age, religion, education, and urbanization. Respondents were approached with the assistance of the Youth Harvest Foundation Ghana (YHFG), churches, mosques, key local figures in the municipality, and the Ghanaian host family of the first author. Respondents were included until data saturation was reached. Data Collection Data were collected in various rounds in the period 2010-2012. In the first stage of the research project (2010-2011), semi-structured interviews that addressed the sexual and relational behavior of the youth were conducted with young people and adults (Van der Geugten et al. 2013). From this first research stage, 12 interviews with 14 young people (two in same-sex pairs) and 17 interviews with adults (six females, 11 males) were selected for the present paper, as they contained the topic ''protected or unprotected premarital sex'' (see Table 2); a secondary analysis was subsequently applied. In 2011 and 2012, 16 semi-structured interviews with 18 young males and females (14 individual interviews, two in same-sex pairs) and five focus group interviews with 39 young males and females were conducted (see Table 2). The main focus was on unprotected and protected premarital sex and on having multiple sexual partners (the latter issue is addressed in another paper; author). The topic list for the semi-structured and the focus group interviews was based on literature and previous research (Van der Geugten et al. 2013). It contained the following five topics: (1) Prevention of STIs; (2) Opinions, conceptions, and motives concerning condom use; (3) Negotiating space for condom use; (4) Expectations of young males and females concerning condom use;

(5) Opinions, conceptions and motives regarding other contraceptive methods. 1 The order in which the topics were discussed in each interview depended on the answers of the participants. The semi-structured interviews lasted 20-75 min, the focus groups 30-60 min. All interviews were digitally recorded and transcribed verbatim. Most interviews with the young males and females and adult respondents were conducted by the first author (Dutch woman, familiar with the research area since 2000) and in the English language. Under the supervision of the first author, four interviews with young people were carried out by a female Dutch undergraduate. Three interviews and two focus group interviews with young people were conducted by a male Ghanaian bachelor graduate. He interviewed only young males, and used Gurune (the local language) in four of the five interviews (without an interpreter) in order to allow young males who did not speak English to be included. A local female interpreter (aged 22) assisted in one focus group interview with young females and in 10 semi-structured interviews with young males and females. These interviews were partly in English and partly in Gurune. Both the interpreter and the Ghanaian interviewer were well known by the first author, and had been selected based on their suitability for this task. Of the interviews with the adult respondents, 16 were held in English and one was held in Gurune with the assistance of a local female interpreter (aged 31). Eleven of the 17 interviews with adult respondents were carried out by the first author; two female Dutch undergraduates carried out six interviews, supervised by the first author. During the various periods of fieldwork, the first author and the Dutch students were hosted by Ghanaian families. This allowed the researchers to experience life in the area and to better understand the social and cultural context. 1 The topic list also included topics concerning multiple sexual partners, which are reported in another paper. Ethical Approval Ethical approval was not needed for this study under Dutch and Ghanaian law during data collection (2010)(2011)(2012). However, the Ghana Health Service and the Navrongo Health Research Center (NHRC) were officially informed and consulted about the project. The research proposal was also discussed with and approved by the YHFG (partner organization in Bolgatanga providing SRH education to young people) and various local authorities. All interviewed persons were informed about the research objectives by the interviewers, asked to cooperate voluntarily and anonymously, and asked orally for their consent. The majority of those approached agreed to cooperate. Some refused because they were not interested or due to obligations at school, home, or work. Three respondents were aged under 16 years; for one of them permission was also asked and given by her mother, the other two participated in the interviews as part of their attendance in a SRH programme for which permission was given by their parents. Respondents could discontinue their cooperation at any time and personal information was excluded from data analysis and publication. Data Analysis The qualitative data analysis software NVivo 10 was used. As a first step, all interviews were coded with the focus on factors that influence unprotected and protected premarital sex, and particularly condom use. Five categories were defined on the basis of these codes. The first author carried out the coding in NVivo 10. Methodological aspects of the research, the coding processes (development of codes and categories), and contradictions that were identified during the analysis were documented and systematically discussed by the research group (JvdG, BvM, MdU, NdV). For privacy reasons, respondents were given fictitious names. Results This section presents the demographics of the respondents and elaborates the five categories that were derived from the data to answer the two research questions. These categories-namely (1) the influence of culture and religion, (2) knowledge of contraceptives, STIs, and pregnancy, (3) attitude and behavior regarding buying and possessing condoms, (4) attitude regarding using condoms, and (5) the influence of unequal power relations between young males and females-provide insight into the factors that influence the attitudes toward unprotected and protected premarital sex of the youth in general and condom use in particular. Demographics Semi-structured interviews were held with 32 young people aged 14-25 years. Eighteen were male (mean age 21.4) and 14 were female (mean age 18.9; one was pregnant), and they differed in age, religion, ethnicity, and education. All respondents were unmarried. Of the males, 11 were sexually experienced and two were not (for five males this was unknown). Of the females, four were sexually experienced and six were not (for four females this was unknown). Sample characteristics are summarized in Table 3. In addition, 39 young people participated in five focus group interviews. Three focus group interviews were held with 22 young males aged 16-25 years (mean age 21.8). Although the age range was announced when selecting the young respondents, three unmarried men aged 29, 30, and 32 years, respectively, were also present. Because the interview had already started when they mentioned their ages, it was thought that it would disturb the group if they were sent away. They were therefore included in the research. One of the focus groups with young males was held in a rural area; participants were either school dropouts, attending junior high school, cowherds, or farmers. The other two focus groups with young males were held in urban areas; most of these participants were attending senior high school (SHS). Two focus groups were conducted with 14 young females aged 16-21 years (mean age 18.1). One was held in a rural area, with six young females who were either school dropouts or attending junior high school (JHS). The other focus group was held with eight young females, all of whom were attending a boarding SHS. The majority of the respondents in all focus groups were Christian; the others were Muslim or Traditionalist. Six female and 11 male adults were also interviewed; their ages, religions, and backgrounds varied. Sample characteristics are summarized in Table 4. The Influence of Culture and Religion In three focus groups, and in four interviews with young females, the respondents spoke about the traditional and religious ideology of abstinence from premarital sex, particularly in relation to the prevention of STIs and pregnancies. It was explained that abstinence is an important way to avoid unprotected sex and its adverse consequences. For example, Gifty (18, SHS student, Christian) said that she believes that, as a virgin, she will not only get the right partner and the respect of the community, but also be protected from STIs and unintended pregnancy. One male (focus group urban area) also said: ''But those who prevent it, to have sex, they prevent sickness and at the same time pregnancy.'' It is tradition among the Frafra in Bolgatanga municipality for fathers to talk to their sons, and for mothers to talk to their daughters, about puberty and its physical changes, abstinence from premarital sex, and the duties of a husband or wife. These talks are initiated the moment a child undergoes physical changes related to physical maturity (e.g., menarche or pubic hair growth). Most uneducated parents still give these traditional talks, but they often do not have enough knowledge of SRH and the risks involved for their children, according to Patrick (social worker) and Sayida (community nurse). In Patrick's words: The other category of parents are the illiterate parents who have no exposure to anything of development and new things of our world, our developing A few young people said that their parents had warned them that unprotected sex can have unintended and unwanted consequences. However, Sayida (community nurse) said that only a relatively small proportion of the educated parents actually talk about SRH to their children, while others send them to the health clinic for SRH education. Apart from the parents' knowledge of SRH, talking about SRH-related issues to young people is still a cultural taboo, according to Patrick (social worker), Sayida (community nurse), John (teacher), and Felix (parent). It is seen as immoral and it is feared that it could encourage young people to have sex. Sayida and Ruth (social worker) taught young people about condom use, and were consequently accused of encouraging immoral behavior. Sayida remarked: In the interviews, Islamic and Christian leaders said that in the church and the mosque they continue to promote abstinence to the youth and discourage condom use. Patrick (social worker) stated that some of the Christian youth ignored the advice of church leaders and used condoms-some did so secretly-to protect themselves against HIV in particular. George (21, Christian, completed SHS) said that he engaged in premarital sex and advocated condom use: ''So if they cannot abstain, then it will be very good for them to use the condom, because you know the abstinence is very difficult for them to do that.'' Some of the adults said that parents have a responsibility for the attitude and behavior of the youth regarding unprotected premarital sex, and their knowledge or lack thereof of STIs, contraceptives, and pregnancy. For example, Sayida (community nurse) said: ''So it's high time we come out openly to talk to our children about sex and about condoms, because they are becoming pregnant.'' Felix (parent) confirmed the taboo on talking about sex, but also said that the traditional ideas are losing ground: Because we think that sex is secret! Sex should not be mentioned. For example, if I talk about penis, they say I'm immoral. (…) So in Africa, that is the problem. But it's, we are overcoming it. It's changing fast. Furthermore, the way parents and children communicate has changed because of increasing modernity, according to one young male and one young female and various adults: Young people nowadays do not obey their parents, they go to friends rather than their parents with their problems, and ''they do what they want.'' Felicity (teacher) said that parents found it difficult to control their children. Francis (25, uneducated farmer) said: The problem with us the youths of these days is the influence from our own peers. It is our friends. For instance, a young boy or girl may have a problem and instead of talking it out with his or her mother or father so that they can give them good advice on how to solve that problem, they would rather talk it out with their friends who most often all they can give is bad advice. It was noticed, however, that in general it was not common in the Frafra tradition for parents and children to converse extensively, or for children to approach their parents with questions. Sayida (community nurse) said: ''Really, do you see a man and a woman and the children conversing? No, in our tradition, when grownups are there, children don't go near there.'' Mohammed (parent and Islamic leader) commented that most parents are ignorant about their children's sexual and relational behavior, and do not know whether they use family planning methods. He also said-as did George (21, Christian, completed SHS, sexually experienced) and Joyce (15, Christian, SHS student, not sexually experienced)-that the youth hide their sexual and relational activities from their parents. Joyce: But the parents will never know. It's when the pregnancy comes, that you run to your parents. That's the only time they will know that your child is having a boyfriend. Because no parent will support that [having a boyfriend]. Unless you are maybe at the age of marriage. Moreover, Ruth (social worker) and John (teacher) said that that it would be better if parents were to talk with their children, they ''can do a lot.'' Ruth said that she wishes parents would make time and create opportunities for their children to share their problems with them. Knowledge of Contraceptives, STIs, and Pregnancy The lack of knowledge of a considerable number of respondents about contraceptives, STIs, and pregnancy came to the fore as a factor that influences the protected or unprotected premarital sexual behavior of the youth, and particularly their condom use. The consequences of unprotected sex, and how STIs (including HIV) are transmitted, were not well known among the youth, according to two young males and two young females. For example, Ayine (21, completed SHS, Traditionalist) said: ''They [youths] are ignorant because they don't know the consequences of having unsafe sex. They just go ahead and go

on to have sex without.'' Abdul (21, SHS student, Muslim) gave the following statement, which he claimed to have learned from a teacher: ''So, if only the girl has HIV/AIDS and she doesn't have a cut, it's not easy for her to give you HIV/AIDS.'' Regarding contraceptives, most respondents knew where to buy male condoms in Bolgatanga municipality, such as pharmacies, supermarkets, and health clinics. However, not all respondents seemed sure about the safety and usage of contraceptives. For instance, Caroline (18, completed JHS, Traditionalist) asked during the interview: ''They always say a condom is not 100% safe. Is it true?'' Several respondents, in a focus group and in the individual interviews, also doubted the safety of male condoms: They wondered whether condoms could burst and whether they could be punctured. It was mentioned in the focus group with female SHS students that the female condom is not safe either. The female SHS students in the focus group and two young females in the individual interviews also expressed concerns that contraceptive pills and injections could cause infertility. Insufficient knowledge was also found with respect to pregnancy prevention. To prevent pregnancies, some young people practice the withdrawal method and some of the young females keep track of their ''safe'' days within their menstruation cycle, without knowing the risks of these methods. In the focus group with the female SHS students, one student said that they could have unprotected sex as long as the young males say that they will not ejaculate inside them. Joyce (15, SHS student, Christian) said about unprotected sex during ''safe'' days: ''Maybe, they will say, well I just had my menses. They think that, oh, I had my menses some time ago, so I don't think if I do it right now I'll get pregnant.'' During the interviews, most young people showed interest in learning more about SRH, the risks associated with unprotected sex, and the use of contraceptives. Ayine (21, completed SHS, Traditionalist) said that education might ''help'' the youth to have protected sex. Particularly education in how to use condoms was mentioned by some of the respondents. One female participant (Christian) in the focus group in a rural area said that ''We need to be educated in how to use condoms, which is very important to us,'' and Clement (20, completed SHS, Christian) said ''Even when you learn that condom is the best way to prevent it, you can go and buy the condom. Yet you don't know how to do it, and it will be useless.'' Sayida (community nurse) confirmed that young people lack knowledge of SRH, and that more females than males come to her with questions and problems regarding STIs and pregnancy. Attitude and Behavior Regarding Buying and Possessing Condoms The attitude and behavior of the youth regarding buying and possessing condoms appears to influence their protected or unprotected premarital sexual behavior, and particularly their use of condoms. A small proportion of the males did not have a problem with going into a shop and buying condoms. George (21, sexually experienced) said: ''So I always feel free, I have one drug store that I always buy, so even sometimes when I get there and I place the money on the table, he knows what I want.'' Three young females also said that they would feel comfortable about buying condoms; one of them confirmed she actually bought condoms. Diana (23, Christian, apprentice seamstress): Me like this, I don't have fear of anything, if I want to, if I'm, my boyfriend, I will always, I will let him know that me I will do it. Because he did not marry me yet. And you know boys, they go out and they come in, they go out and they get different, different ladies, ahaa. So it's only good. Most young males and females, however, said that they felt uncomfortable about buying condoms. One female SHS student said in a focus group: I think due to that too, like the shyness, I may not go and buy it myself. So, I will say, maybe I feel shy to go to the drugstore and tell the person that I want condom or something like that. Some of the young males also experienced this barrier. Hashim (23, completed SHS, Muslim, sexually experienced): Personally me like this, I feel so shy that I feel heavy to go there to buy a condom (…). You see that like the traditional teaching in Bolga here, is that already sex before marriage is a crime. We see it as something that is odd.(…) It is rampant. It is everywhere but still the elders see it (…) It is a bad behavior. In the focus group with female SHS students and in the interview with Hashim, it was said that young people could also ask a younger person (it is culturally accepted to assign younger persons, especially younger siblings), or an older brother or friend, to buy condoms for them if they themselves felt too shy to go. Sophia (22) and Claudia (14) mentioned the use of different names to hide their order for a condom such as ''cd,'' ''doncom,'' and ''this thing.'' Although most young males and females said they felt uncomfortable about buying and possessing condoms, this feeling seemed to be stronger in females than in males. This difference in attitude was brought up in both focus groups with young females, and in the interviews with four young females and two young males (various religions, aged 14-23 years). According to them, buying and possessing condoms can lead to remarks such as ''bad girl,'' ''bad boy,'' or ''you are too young,'' and that young females would be seen as prostitutes. Ruth (social worker) also said that, in general, it was socially more acceptable and easier for males to buy and possess condoms compared to females. Regarding the possession of condoms, some young males and females said that parents do not allow their sons or their daughters to possess condoms, because they should abstain from premarital sex. Attitude Toward the Use of Condoms Another factor that influences whether young people have protected or unprotected premarital sex, is their attitude toward the actual use of condoms. Almost half of the young females and three of the young males (14-23 years, various religions, all educated) said that it is important to protect oneself against STIs and pregnancies by using a condom. Rudolf (24, Traditionalist, completed SHS, sexually experienced) said that it is important to have a condom available: ''It's not difficult, so far there is any time that you come into my room, I will make sure the condom is always available before having the sex. Don't do it without condom.'' Albert (20, Christian, completed SHS, sexually experienced) also explained that his future, and in particular his education, motivates him to use condoms: But to me, I believe in using condoms because, you know (…) it's better like, to keep your feelings okay. But I believe in my future more than in sex, so for me going to the high level is better than to have sex and maybe contract any diseases. However, according to the majority of the respondents (various ages, religions, and educational backgrounds), not all young males and females want to use condoms. In a focus group with young males in an urban area, one participant said: Some boys prevent it through using condoms, but some boys don't prevent. They don't even prefer using condom. When you ask them why, they say ''Oh, when they use the condom they don't feel.'' They don't feel like making love. Gregory (24, completed SHS, rural area, sexually experienced) said the following about what young males and females feel about using condoms: Yes, because there are some girls or some boys, that if they are using the condom to have the sex, they don't feel like, they don't feel the percentage they are supposed to, like you see, using a condom to have sex with, or using without condom, the percentage is always different. It's different (…) there is much feeling if you use without condom, and there is less feeling if you use with condom. So that's why some boys and there are some girls that if you even use condom with them they won't like. Saïda (23, completed SHS, Muslim, rural area, sexually experienced) made a negative statement about condom use: ''Some too are there with the decision that they don't want to use it. Simply because they always say they feel when they use it for sex, they don't really enjoy much of the sex.'' However, George (21, Christian, completed SHS, SRH peer educator, sexually experienced) held a different opinion: ''It's the same thing, whether condom or not, it's the same feelings, it's the same, what will happen will still happen.'' He also said that there are girls who do not like the use of condoms because they can cause heat friction in their vaginas. One of the social workers (26, female, Christian), who was also an SRH educator, remarked that most young people preferred not to use condoms: So if you tell them that, put the condom on before you have sex. They are, like, when you buy a toffee, do you just put it in your mouth like that? (…) So you see they are trying to tell us that, if you put a candy in your mouth like that you don't, uhmm, you don't feel the taste immediately, you have to chew it or something when you come. When they use condom, they don't get satisfied, or they don't get the feeling. So they prefer to go the natural way. They call it the raw way. It was noted that peers, especially same-sex peers, have an important role in advising and influencing each other regarding SRH. Although most young people said that sexual intercourse with a condom gives less pleasure and less ''feeling'' than without a condom, most of them did not talk about their own experience, but about what they had heard from their peers. For example, Thomas (21, Christian, completed SHS, not sexual experienced): I have enough experience from my friends. That when you use protection with girls it's like you don't feel. You don't feel. (…) Even though sometimes they tell me but I've never been in their shoes before. I've never done such thing. So I believe them. 'Cause they have been doing it. In two interviews and one focus group with young females, it was said that males sometimes cut off half or the top of the condom, without their partners realizing it, in order to ''have more feeling.'' One female SHS student (Christian) said in the focus group: ''Sometimes they will pretend to use it, meanwhile they will remove it. They can intentionally cut half of it.'' Mutual trust between boyfriend and girlfriend is an important factor that influences the use of condoms, according to three young males and one young female. When young people trust each other they do not use protection, and in the case of distrust, they like to use a condom. Suggesting using a condom could therefore make a partner suspicious: The one suggesting it might have an STI and not admit it. David (20, Christian, university student) said that young people are afraid their partner would break up the relationship if they were honest about having an STI. In the focus group with the female SHS students, it was mentioned that young males want their first sexual intercourse to be ''raw'' (without a condom): ''Because the first time they have sex with you, they want to do that so that you know that you are getting to like him or not.'' Additionally, Francis (25, uneducated farmer) said that it is important to find out soon whether you are ''sexually compatible'': ''It is very necessary for you to try her fast and see how she tastes.'' Deciding to use condoms after being tested positive for HIV or another STI was mentioned in a focus group with young males in a rural area, and by three young males and two young females in individual interviews. However, none of them reported having actually had such a blood test. Unequal Power Relations Between Boys and Girls It was found that unequal

power in sexual relationships between young males and females is also important when it comes to the actual use of condoms. Three young females said that if males insisted on not using a condom, females could not force them to do so. Saïda (23, completed SHS, Muslim, rural area) pointed out that young males are physically stronger than young females: It's not easy, if the guy doesn't want it, I don't think you can find it easy using it. Because you cannot force him. And mostly the guys have the power over the ladies, that they can force them to do what they want, but you a lady, simply because our strength is not equal. Another reason given for agreeing to have sex without a condom, is the girl's love for her boyfriend. Samira (21, SHS student, Muslim, rural area): ''If the boy says 'I don't like using condom,' and then the girl too thinks that she loves the boy, you see that she will just give herself to him. Without a condom.'' A female youth's financial dependence on her boyfriend might also contribute to unprotected sex. According to three young males and three young females and some of the adults, some of the young females in the research area have transactional, unprotected sex with young males in order to provide for their needs. It allows them to buy the food, clothes, and luxury items that their parents are too poor to provide. Some of the young females were encouraged by their female friends to engage in these transactional sexual relationships, despite the health risks. Caroline (18, completed JHS, Traditionalist) said: Yes, because when you are about two, three or four girls going, you see that one of the girls can go into a boy, that is giving her much money, or plenty money. And the fellow [girl] can influence you that you also go and friend this guy, he's having money, he also gives you the money. Not knowing that that guy is having these diseases. And you too go into that guy, and he gives you money. After giving you the money he also wants to get something from you. Condom use was seen as the males' responsibility, according to one young male and two young females. They said that young males are mostly in charge of buying condoms and having them handy, that they should initiate condom use, and that they are the ones who have to wear the condom. Lydia ([20, completed university, Christian) said: ''It's like the guy always uses the condom so they [young females] don't really have much to say about that.'' Rudolf (24, Traditionalist, completed SHS) said: And then the selling of condoms is now open to any place, you can go to any shop and then you get the condom to buy. (…) Yeah they normally, mostly it's the boys' one that is common in our community here. However, not all girls accepted the unequal power relation between males and females. One sexually experienced female youth said that the use of condoms was her responsibility (17, JHS student, Traditionalist, rural area). She said that she would insist on condom use, even if the boy did not agree: ''Yes, if he will not agree then I will stop it, I will say I will not do it. And he cannot force you to do it.'' Discussion This study focused on conceptions and attitudes of the youth toward protected and unprotected premarital sex, and in particular condom use, in Bolgatanga municipality, Ghana. Various reasons and motivations for these sexual practices were identified: (1) the influence of culture and religion, (2) knowledge of contraceptives, STIs, and pregnancy, (3) attitudes and behavior regarding buying and possessing condoms, (4) attitudes regarding using condoms, and (5) the unequal power relations between young males and females. Young peoples' lack of comprehensive knowledge of STIs, contraceptives, and pregnancy, as well as their ambivalent and inconsistent attitude regarding condoms and other contraceptives, is probably rooted in the taboo on premarital sexuality, which is related to the strong promotion of premarital sexual abstinence by the Traditionalist, Christian, and Islamic religions in the research area. Not all parents, schools, churches, mosques, and organizations educate young people comprehensively about SRH, and some even discourage SRH education because they fear it would encourage young people to have sex. Parents' reluctance to teach their children about SRH and limited communication between parents and children were both observed in the present study. Moreover, it was noticed that parents themselves had a lack of knowledge regarding SRH. It was also found that there is a gap between traditional cultural and religious values on the one hand, and increasing modernity and education on the other hand, which creates a distance between parents and children. The growing influence of Christianity and Islam, which promote abstinence from premarital sex, strengthens the conviction held by parents that premarital sex education is unnecessary. The norms to abstain from premarital sex and the taboo on educating young people about SRH were easy to maintain in the past, when young people mostly remained within their communities until they were married. Nowadays, most young people have more independent and autonomous lives: They visit their friends in other communities, use the internet on smart phones, and go to school or to work. In order to have a more effective policy regarding premarital sexual behavior, parents, as well as schools and religious organizations, need to adjust to these developments. This requires a different approach, one that may require parents, teachers, and other key figures to increase their knowledge regarding: (1) Sexual and reproductive health and rights in general and particularly contraceptives; (2) the positive effects of early SRH education; (3) places where young people can obtain SRH education or consult an SRH professional; and (4) the adverse consequences of unprotected sex. They might also have to improve their skills in discussing SRHrelated issues with young people. Research in other sub-Saharan African countries has shown that communication about SRH between children and parents is not common (Biddlecom et al. 2009), and that programs can help parents to improve this communication, which has positive effects on the adolescents' health (WHO 2007). As stated, national Ghanaian data show that 11% of unmarried sexual active girls used condoms in 2014, and that less than 40% of boys aged 15-19 used condoms in 2008 (Ghana Statistical Service, Ghana Health Service andICF International 2015;Doyle et al. 2012). A recent study among junior high school students in Bolgatanga (Krugu et al. 2016b). We recommend the execution of more quantitative research on this topic. It is a promising finding in the current study that almost half of the girls and some of the boys had a positive attitude toward using male condoms to prevent STIs and pregnancy. However, the majority were uncomfortable about buying and possessing condoms, which might be related to the cultural and religious taboo on premarital sex. Other contraceptives-such as pills, injections, and female condoms-were hardly used or mentioned by respondents in the present study. Moreover, the safety of these other contraceptives was questioned by some of the girls, who feared they can cause infertility. Concerns of infertility when using contraceptives were also found among women in Accra (Hindin et al. 2014). It is worrisome that young people still do not have easy access to contraceptives, or feel uncomfortable about obtaining or using them, particularly since the Ghanaian government has been promoting condom use for more than 10 years now, and research as long ago as 1997 showed that the Ghanaian youth feel embarrassed about buying and carrying condoms (Awusabo-Asare et al. 2004). It was found that same-sex peers play an important role with respect to the SRH of the youth in Bolgatanga municipality: Most of the young respondents in this study did not talk about their personal experience with condoms, but mainly about what they had heard from their peers. Peers generally share stories and inform each other about SRH issues, but the ''facts'' they communicate are not always correct or are merely personal preferences. The fact that most of the youth do not talk about their personal experience can be attributed to various factors, for example, their lack of sexual experience, the cultural and religious taboo, or their personal shyness related to this taboo. In order to improve SRH education, it is important to include the influence of peers in SRH education, particularly because tales of unpleasant experiences with contraceptives spread rapidly among peers, which was also found in a small study in Accra, the Ghanaian capital (Appiah-Agyekum and Kayi 2013). The use of condoms might cause suspicion and distrust between sexually active young people, because it might lead them to doubt each other's faithfulness. This suspicion might also be influenced by the ''ABC'' strategy that the Ghanaian government has been promoting for over a decade: If one cannot abstain (A) from sex or be faithful (B), then use condoms (C). Using a condom might thus imply that one has other sex partners as well. The prevalence of polygyny in Bolgatanga municipality (in 2011, it was practiced by 25% of men and 39% of women (Ghana Statistical Service 2012) might also contribute to distrust: The acceptance of polygyny coupled with the unlimited sexual freedom of Ghanaian men inside and outside marriage (Anarfi and Owusu 2011), means there will always be suspicion among women, even when their partners declare that they are faithful. The unequal power balance in premarital sexual relationships between young males and females-which comes to the fore in explanations about physical strength, being helplessly in love, or female financial dependence-also plays a role in whether young people have protected or unprotected sex. It was said that because males are physically stronger, it is difficult for females to insist on condom use. Moreover, condoms were mostly seen as the responsibility of males: The condom is a ''male'' prophylactic, and it is embedded in a culture in which males have more authority than females. The findings that young males and females have limited knowledge of SRH and contraceptives, and that condom use is seen as unpleasant and might cause distrust in sexual relationships, are in accordance with previous research in other sub-Saharan countries and northern parts of Ghana (Kuumuori Ganle et al. 2012;Rondini and Kingsley Krugu 2009;Sayles et al. 2006). The qualitative interviews with young people and adults in the present study-which revealed young peoples' feelings, fears, shyness, and dilemmas regarding protected and unprotected sex, condom use, and premarital sexual abstinence-contribute to a better understanding of their premarital sexual behavior. Understanding the influencing factors that motivate young males and females to have protected or unprotected sex, and understanding these factors in specific contexts, can contribute to the development of more tailored and effective SRH education to protect young people from the adverse consequences of engaging in unprotected sex (Darteh et al. 2014;Karim et al. 2003;Madise et al. 2007). Strengths and Limitations A strength of this study is that young people in a remote region in northern Ghana shared their perceptions of sexual relationships, something that has rarely been done before. Another strength is that qualitative research was used, which provides indepth insights into and information about the research topic. Additionally, using three interviewers increased the credibility of the results. Two Dutch females interviewed both males and females, and they were looked upon as ''outsiders'' who would leave again. This ensured the privacy of the respondents, which is particularly important for young females, who are supposed to be virgins. One Ghanaian male interviewed only young males, in order to decrease possible bias caused by only females interviewing young males. The validity of this study was increased by using two data collection methods: individual and focus group interviews. Addressing personal topics in a focus group interview could be a limitation, particularly when respondents do not feel confident in a group. In this study, it was experienced that young people were motivated to react on each other and shared their point of view. Although the individual interviews gained more indepth stories of respondents, the focus group interviews were relevant complementary data. This study also had its limitations. The results might have been influenced by the fact that the study included among the young males and females more older respondents (C18) than

younger respondents (B17), and relatively more Christians than Muslims and Traditionalists. Further, data on religion and ethnicity were missing for some of the respondents. Finally, an interpreter was used in some of the interviews, which could have influenced the conversation, and for practical reasons, member checks (respondents checking the interview transcripts) could not be done. Implications In the development and delivery of SRH programs, it should be taken into account that the traditional and religious ideas concerning premarital sex contradict the modern, western-oriented ideas in Bolgatanga municipality. Both streams of ideas influence the knowledge of contraceptives, STIs, and pregnancy of the youth, which turns out to be limited and sometimes also incoherent. Although the attitudes of young people toward the use of male condoms to prevent STIs and pregnancy are in general positive, there are various reasons why they do not use them consistently. SRH education should address the misinterpretations regarding the safety and side effects of contraceptives, as well as how to deal with distrust in sexual relationships and how to handle the stories told by their peers. Additionally, young males and females should be enabled to access contraceptives easily and without feeling uncomfortable or stigmatized. It is important to involve parents and other key figures in SRH education. The traditional family role of preparing young people for marriage and informing them about SRH issues needs to change now that an increased number of young males and females have more autonomous lives, go to school, and have premarital sex. For instance, teachers, religious leaders, and health and social workers could promote the healthy sexual and reproductive development of young people by providing comprehensive SRH education or counseling. Additionally, the increased use of the internet on smart phones by young people could contribute to improve their knowledge of SRH, but only when they are taught how and where to find reliable information. Compliance with Ethical Standards Conflict of interest All authors declare that they have no conflict of interest. Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed Consent Informed consent was obtained orally from all individual participants included in the study. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Structure of the Colony Migration Factor from Pathogenic Proteus mirabilis A CAPSULAR POLYSACCHARIDE THAT FACILITATES SWARMING* Swarming by Proteus mirabilis is characterized by cycles of rapid and coordinated population migration across surfaces following differentiation of vegetative cells into elongated hyperflagellated swarm cells. It has been shown that surface colony expansion by the swarm cell population is facilitated by a colony migration factor (Cmf), a capsular polysaccharide (CPS) that also contributes to the uropathogenicity of P. mirabilis (Gygi, D., Rahman, M. M., Lai, H.-C., Carlson, R., Guard-Petter, J., and Hughes, C. (1995) Mol. Microbiol. 17, 1167–1175). In this report, the Cmf-CPS was extracted with hot water, precipitated with ethanol, and further purified by gel permeation chromatography. Its structure was established by glycosyl composition and linkage analyses, and by oneand two-dimensional NMR spectroscopy. The Cmf-CPS is composed of the following tetrasaccharide repeating unit. Proteus mirabilis is a pathogenic Gram-negative bacterium that frequently causes kidney infections, typically established by ascending colonization of the urinary tract (1)(2)(3)(4)(5). It exhibits a striking form of multicellular behavior, called swarming migration, in which motile vegetative rods growing on solid media differentiate into extremely elongated hyperflagellated swarm cells that undergo rapid and coordinated population migration away from the initial colony (4,6,7). A transposon mutant of P. mirabilis, WT19, defective in mass migration of normally differentiated swarm cells, has been reported (8). Genetic analyses identified a lesion in a putative polysaccharide assembly locus, and electron microscopy and gel electrophoresis confirmed the specific loss of a capsular polysaccharide (CPS). 1 This CPS, named Cmf (colony migration factor)-CPS, was suggested to facilitate population migration by enhancing medium surface fluidity and possibly influencing cell-cell interactions. The cmfA Ϫ was also attenuated in experimental uropathogenicity, showing greatly reduced colonization of the urinary tract (9). Little is known about the structures of Proteus polysaccharides, and serology indicates substantial heterogeneity, with P. mirabilis and the closely related Proteus vulgaris divided into 49 O-serogroups, and many smooth strains remain unclassified (10). Analysis of the Cmf-CPS from wild-type P. mirabilis WT19 (8) indicated that it is an acidic type II molecule rich in galacturonic acid and N-acetylgalactosamine and that it can have a phospholipid anchor. This composition showed that it must be structurally different from previously reported, functionally anonomous CPSs from P. mirabilis ATCC49565 (11) and P. vulgaris CP2-96 (12). The former was reported to consist of a branched trisaccharide repeating unit of N-acetylglucosamine, N-acetylfucosamine, and glucuronic acid and the latter of a tetrasaccharide repeating unit of two glucosyl, one N-acetylgalactosaminosyl, and one glucuronosyl residue. To establish the nature of a functionally characterized Proteus polysaccharide and gain a view of the possible common structural features among the polysaccharides of this genus, we report the structure of the Cmf-CPS from P. mirabilis WT19. EXPERIMENTAL PROCEDURES Bacterial Strains and Growth Conditions-P. mirabilis WT19 strain U6450 (proticine type P3/S1) (13) was isolated from a chronic urinary tract infection involving renal stone formation (13). The bacterial cells were grown overnight at 37°C on the surface of Brilliant Green agar (BBL, Becton Dickinson, Cockeysville, MD), and colony morphology was examined the next day. After observing that this culture produced terraced, swarming colonies that extended across 50% of the agar surface, cells at the edge of the swarm colony were transferred to Brilliant Green agar again to confirm the absence of contaminants. Biochemical confirmation of this strain as Proteus was made using a commercial package of diagnostic reagents (Enterotube II, Becton-Dickinson) and was confirmed as P. mirabilis by the National Veterinary Services Laboratories (Ames, IA), although it atypically failed to ferment two sugars, maltose and xylose. Isolation and Purification of Cmf-CPS-Bacteria were grown in BHI broth, harvested by centrifugation, and washed once in physiologically buffered saline as described by Lee and Cherniak (14). Bacterial cells (100 g, wet weight) were suspended in 300 ml of water and stirred vigorously in boiling water for 30 min. The suspension was cooled in an ice bath with stirring for 90 min. The cell residue was removed by centrifugation (10,000 ϫ g, 30 min, 4°C), the supernatant adjusted to 1% acetic acid, and crude Cmf-CPS precipitated by the addition of ethanol (2.5 volumes, 24 h, Ϫ20°C). The Cmf-CPS precipitate was collected by centrifugation (10,000 ϫ g, 30 min, 4°C), washed with ethanol, washed again with acetone, dried, dissolved in water, and lyophilized. This crude Cmf-CPS was suspended in a solution containing 6 ml of EDTA-phosphate (0.05 M Na 2 HPO 4 /0.005 M EDTA, pH 7.0), 3 mg of DNase (in 3 ml of 0.04 M MgCl 2 ), and 20 mg of RNase (in 3 ml of 0.04 M MgCl 2 ). This solution was incubated for 16 h at 37°C followed by the addition of proteinase K (4 g) and incubated again for 16 h at 37°C. The resulting solution was dialyzed against distilled water for 48 h and centrifuged at 5,000 ϫ g for 20 min, and the supernatant was lyophilized. The final yield was 800 mg of crude Cmf-CPS. Crude Cmf-CPS was further purified by gel filtration column (90 ϫ 1.6 cm) chromatography using Sephadex G-150 equilibrated with a buffer solution consisting of 0.2 M NaCl, 1 mM EDTA, 50 mM Tris base, and 0.25% deoxycholic acid (DOC), pH 9.25. The content of each fraction was identified by polyacrylamide gel electrophoresis in the presence of DOC (DOC-PAGE) using 18% acrylamide (15). Gels were fixed in the presence or absence of Alcian blue (16) and silver-stained (17). A fraction of the crude Cmf-CPS was also further purified by mild acid hydrolysis in aqueous 1% acetic acid at 100°C for 2 h. After hydrolysis, the solution was cooled and centrifuged (10,000 ϫ g). The supernatant was extracted with diethyl ether (3 ϫ 10 ml), and the aqueous layer was fractionated on a Sephadex G-75 column (90 ϫ 1.6 cm). The fractions were assayed for hexose with phenol-sulfuric acid. The resulting Cmf-CPS and oligosaccharide (OS) fractions were lyophilized. Nuclear Magnetic Resonance Spectroscopy-Samples were prepared for NMR analysis by a 2-fold lyophilization from D 2 O and dissolved in 0.5 ml of D 2 O. Spectra were recorded at 60°C. Chemical shifts are reported in ppm, using sodium 3-trimethylsilylpropanoate-d 4 (␦ H 0.00), and acetone (␦ C 31.00) as internal references. All NMR spectra were recorded on Bruker AMX-500 or DRX-600 MHz spectrometers. Twodimensional DQF-COSY (18), TOCSY (19,20), and NOESY (21) data sets were collected in phase-sensitive modes using the States-TPPI (22) method. In these experiments, low power presaturation was applied to the residual water (HOD) signal. Typically, data sets of 2048 (t 2 ) ϫ 512 (t 1 ) complex points were collected with 16 scans per FID and a sweep width in both dimensions of 6 ppm. The TOCSY experiments contained MLEV17 (23) mixing sequences ranging from 60 to 320 ms, and the NOESY mixing delay was 200 ms. A gradient HSQC (24) data set was collected using the echo/anti-echo method for pure absorption data. A data set of 2048 (t 2 ) ϫ 512 (t 1 ) complex points was acquired, with 32 and 64 scans per FID. The sweep width was 7 ppm for proton (F 2 ) and 60 ppm for carbon (F 1 ). The GARP (25) sequence was used for 13 C decoupling during aquisition. Data were processed typically with a lorentzian-to-gaussian weighting function applied to t 2 and a shifted squared sine bell function and zero-filling applied to t 1 . Data shown were processed with Felix software (Molecular Simulations, Inc.). Glycosyl Composition Analyses-Glycosyl composition of Cmf-CPS (0.5 mg each) was performed by hydrolysis in 0.5 ml of 2 M trifluoroacetic acid in a closed vial at 120°C for 3 h. The glycoses in the hydrolysate were reduced with NaBH 4 , acetylated, and analyzed by combined gas-liquid chromatography mass spectrometry (GLC-MS) (26). For the determination of uronic acid, the Cmf-CPS sample (0.5 mg) was dried in vacuum and methanolyzed in 1 ml of methanolic 2 N HCl at 80°C for 16 h. The resulting methyl glycosides were either trimethylsilylated and analyzed by GLC-MS (26), or they were reduced with NaBH 4 (10 mg) in water (100 l), acetylated, and analyzed by GLC-MS. The absolute configurations of the glycoses present were determined by GLC-MS analysis of the trimethylsilylated (S)-(ϩ)-2-butyl and (S)-(Ϫ)-2-butyl glycosides (27,28). Glycosyl Linkage Analyses-Glycosyl linkage analysis was carried out using a modified NaOH method (29,30). The sample (1 mg) was dissolved in dimethyl sulfoxide (100 l), powdered NaOH (100 mg) was added, and the reaction mixture was stirred rapidly at room temperature for 30 min. Methylation was performed by the sequential additions of methyl iodide (10, 10, and 20 l) at 10-min intervals. After an additional 20 min of stirring, 1 ml of 1 M sodium thiosulfate was added, and the methylated glycans were recovered in the organic phase by extraction with chloroform (0.5 ml ϫ 3). The permethylated product was further purified by reverse-phase chromatography using a Sep-Pak C18 cartridge (31). The methylated glycan was hydrolyzed with 2 M trifluoroacetic acid (120°C, 3 h), reduced with NaB 2 H 4 , acetylated, and analyzed by GLC-MS (26). For the linkage determination of the uronic acid, the permethylated Cmf-CPS sample (0.5 mg) was dried in vacuum and methanolyzed in 1 ml of methanolic 2 N HCl at 80°C for 16 h. Released, partially methylated methyl glycosides were N-acetylated with the addition of 200 l of methanol, 20 lЈof pyridine, and 20 l of acetic anhydride at room temperature for 5 h and dried in air. The carboxyl methyl ester of uronic acid was reduced with NaB 2 H 4 (10 mg) in water (100 l) for 2 h at room temperature, neutralized with acetic acid, and dried in

air with the addition of methanol. The carboxyl reduced products were hydrolyzed with 2 M trifluoroacetic acid (120°C, 3 h), reduced with NaB 2 H 4 , acetylated, and analyzed by GLC-MS (26). GLC-MS analyses were performed using capillary columns (length, 30 m; inner diameter, 0.32 mm) with helium as the carrier. A DB-5 column (J & W Scientific) was used for aminoglycosyl derivatives, and an SP2330 column (Supelco, Bellefonte, PA) was used for the neutral glycosyl derivatives. RESULTS Isolation and Purification of Cmf-CPS-Analysis of the crude Cmf-CPS by DOC-PAGE ( Fig. 1) showed that the crude Cmf-CPS contained some contaminating LPS. The crude Cmf-CPS was separated from the contaminating LPS by a Sephadex G-150 column using buffer containing DOC. The DOC-PAGE analysis of the fractions resulted in a fraction that silverstained only after fixing the gel in the presence of Alcian blue (Fig. 1A), a property characteristic of acidic polysaccharides (32). The low molecular weight fraction silver-stained after being fixed in the presence or absence of Alcian blue, a feature that is characteristic of LPS (32). Thus, the crude Cmf-CPS fraction contained a high molecular weight Cmf-CPS DOC and a low molecular weight LPS (a lipo-oligosaccharide, LOS DOC ). During the purification of Cmf-CPS DOC described in the previous paragraph, the sample was subjected to alkaline conditions (pH 9.25) for an extended period of time. Thus, any O-acetyl substituents, if present, would have been removed. To purify Cmf-CPS without removal of O-acetyl groups, a portion of crude Cmf-CPS (50 mg), was hydrolyzed with mild acid and purified by Sephadex G-75 column chromatography. Two fractions were obtained, the high molecular weight Cmf-CPS G75 and low molecular weight oligosaccharides (OS G75 ) derived from the LOS. The Cmf-CPS G75 eluted just after the void volume, and the OS G75 eluted at twice the void volume (not Table I. Both Cmf-CPS DOC and Cmf-CPS G75 have very similar glycosyl compositions, i.e., mannuronic acid, galacturonic acid, N-acetyl glucosamine, and Nacetyl galactosamine in a molar ratio of 1:1:1:1. The LOS DOC contains glycosyl residues characteristic of LPS core oligosaccharides, i.e., glucose, galactose, mannose, D,D-heptose, and L,D-heptose (33). Fatty acid analysis showed that neither of the Cmf-CPS fractions contained detectable fatty acyl residues, whereas the LOS DOC contained myristic, palmitic, and ␤-hydroxy myristic acids, a result consistent with the presence of lipid A. Determination of the absolute configurations of the glycoses present in the Cmf-CPS fractions revealed that all glycoses had the D-configuration. Glycosyl Linkage Analysis-Glycosyl linkage analysis was performed by methylation followed by hydrolysis, reduction, and preparation of alditol acetates. Linkage analysis of the uronic acid was performed by methanolysis followed by reduction of the permethylated sample prior to hydrolysis. The glycosyl linkages of the Cmf-CPS DOC are shown in Table II. Prior to carboxyl group reduction, 3-linked N-acetylglucosamine (GlcNAc), and 3,4-linked N-acetylgalactosamine (GalNAc) were present in a 1:1 ratio. After the carboxyl group reduction, two additional glycosyl residues were observed. The mass spectra and retention times of their partially methylated alditol acetates were consistent with 3,6-linked mannosyl and 6-linked galactosyl residues with two deuteride atoms at C-6, showing that these two residues were derived from 3-linked mannuronic acid and terminally linked galacturonic acid, respectively. NMR Spectroscopic Analysis-The 1 H NMR spectrum of the Cmf-CPS DOC (Fig. 2) confirmed that galactosamine and glucosamine were N-acetylated, as indicated by a singlet at 2.05 ppm. The 1 H NMR spectrum of Cmf-CPS G75 fraction (data not shown) was identical to that of Cmf-CPS DOC . The absence of a signal at about ␦ 2.10 in either Cmf-CPS fraction indicates that the Cmf-CPS does not contain O-acetyl substituents. The anomeric region shows three downfield ␣-anomeric proton signals (Table III) With the aid of two-dimensional COSY (spectrum not shown), TOCSY (Fig. 3A), and broad-band decoupled HSQC (spectrum not shown) analyses, most of the 1 H and 13 C NMR signals could be assigned (Tables III and IV). The four glycosyl residues were designated A-D according to their decreasing anomeric chemical shifts. Residue A has an anomeric signal at ␦ 5.37 and a J H-1,H-2 coupling constant of 3 Hz, indicating that it is an ␣-linked residue. The H-1 to H-5 proton signals (Table III) for residue A were assigned from the COSY and TOCSY (Fig. 3A) spectra. A large J H-3,H-4 coupling constant (Ͼ5 Hz) was observed for A, supporting the conclusion it has a gluco configuration. The carbon signals (Table IV) from C-1 to C-5 for residue A were determined from the HSQC spectrum. The C-2 chemical shift of residue A is ␦ 54.0, typical of a nitrogen bearing carbon. The downfield position of the C-3 carbon signal (␦ 82.2) indicates that residue A is substituted at this position. Thus, A is the 3-linked N-acetylglucosaminosyl residue. The carbon chemical shifts from C-1 to C-5 for residue A (Table IV) are also similar to those previously reported for a 3-linked ␣-N-acetylglucosaminosyl residue (34). a These values are for the partially methylated alditol acetates from carboxyl-reduced (CR) Cmf-CPS DOC . The mass spectra of these partially methylated alditol acetates show that they both have two deuterium atoms at C-6 and indicate that they were derived from the 3-linked mannuronic acid and from terminally linked galacturonic acid present in the Cmf-CPS. a Mass spectrometric analysis shows that the alditol acetates of these residues have two deuterium atoms at C6, indicating that they were derived from their corresponding uronosyl residues, i.e. mannose and glucose from mannuronic and glucuronic acids, respectively. The anomeric signal for residue B is ␦ 5.17 (J H-1,H-2 not resolved), showing that it is ␣-linked. The proton chemical shifts (Table III) from H-1 to H-5 protons were assigned from COSY (spectrum not shown) and TOCSY (Fig. 3A) spectra. A relatively small J H-3,H-4 coupling constant (Ͻ5 Hz) for residue B indicates that it has a galacto configuration. The carbon chemical shifts (Table IV) from C-1 to C-5 for residue B were assigned from HSQC spectrum. The C-2 chemical shift of res-idue B is ␦ 50.5, typical of a nitrogen bearing carbon. The downfield shift of C-3 (␦ 76.6) and C-4 (␦ 79.1) indicates that residue B is substituted at C-3 and C-4. Glycosyl linkage analysis (Table II) showed that N-acetylgalactosamine is the only 3,4-linked aminoglycosyl residue found in the Cmf-CPS. Therefore, residue B is the 3,4-linked-␣-N-acetylgalactosaminosyl residue. Residue C has an anomeric proton chemical shift at ␦ 5.00, (J H-1,H-2 not resolved) indicating that it is ␣-linked. The proton chemical shifts (Table III) from H-1 to H-5 for residue C were assigned from the COSY and TOCSY (Fig. 3A) spectra. A small J H-2,H-3 coupling constant (Ͻ5 Hz), indicates that the residue C has a manno configuration. The carbon chemical shifts (Table IV) from C-1 to C-5 were assigned from HSQC spectrum. The downfield chemical shift of C-3 (␦ 76.5), indicates that residue C is substituted at this position. Glycosyl linkage analysis (Table II) showed only one 3-linked mannuronosyl residue in the Cmf-CPS. Thus, residue C is the 3-linked mannuronosyl residue. The anomeric proton chemical shift for residue D is ␦ 4.79 (J H-1,H-2 7 Hz) indicating that it is ␤-linked. The proton chemical shifts (Table III) from H-1 to H-5 for residue D were assigned from the COSY and TOCSY (Fig. 3A) spectra. The J H-3,H-4 coupling constant for residue D is similar to that for residue B (i.e., Ͻ 5 Hz), indicating that it has a galacto configuration. The carbon chemical shifts (Table IV) from C-1 to C-5 carbon for residue D were determined from the HSQC spectrum. There is no downfield chemical shift for any carbon of residue D, indicating that it is not substituted. The only terminally linked hexosyl residue observed in the glycosyl linkage analysis of the Cmf-CPS (Table II) was galacturonic acid. Thus, residue D is the terminally linked ␤-galacturonosyl residue. The sequence of glycosyl residues was determined from a NOESY experiment (Fig. 3B and Table V). In addition to intraresidue NOE contacts to H-2 and H-3, residue A has an NOE contact from H-1 to H-3 of residue C. Because residue C is 3-linked ␣-D-mannuronic acid, the sequence shown in Structure 1 was established. The signals labeled in bold type on the NOESY spectrum indicate the strong inter-residue NOE contacts from which the glycosyl sequence was deduced. The mixing time for the TOCSY spectrum shown was 120 ms. Complete assignment required several TOCSY experiments requiring several mixing times ranging from 60 to 320 ms. The spectra for these other experiments are not shown. Residue B has a strong NOE contact from H-1 to H-3 of residue A, indicating that residue B is linked to the 3-position of residue A. Thus, the P. mirabilis WT19 Cmf-CPS consist of a tetrasaccharide repeating unit,n -, as shown in Structure 4. DISCUSSION Extracellular polysaccharides are central to bacterial survival, particularly against the immune defenses of the mammalian host. In uropathogenic P. mirabilis, the Cmf capsular polysaccharide has also been shown to facilitate the rapid multicellular migration of elongated hyperflagellated swarm cells, which correlates with the ability to establish experimental ascending colonization of the urinary tract and may be coupled to the formation of biofilms (7). Proteus migration requires close cell-cell contact, with swarm cells aligning along their long axes in multicellular rafts. The Cmf-CPS may provide a matrix for surface migration of the swarm cell rafts (35), stabilizing cell-cell contact and facilitating intercellular communication (8,35). In addition, the acidic CPS is thought to act as lubricant, creating a fluid environment through which Proteus can swarm by extracting water from the agar medium beneath the colony (4,8). This latter hypothesis is supported by the observation that increased agar concentration or reduced polysaccharide biosynthesis, both of which result in a lowered agar/capsular polysaccharide osmotic activity ratio, reduce migration velocity but do not inhibit differentiation (4,8). Surface active agents are produced by other bacteria that undergo population migration cell rafts, e.g. the unrelated Myxococcus produces an extracellular slime during fruiting body development (36). Increasing the understanding of the apparently multiple roles of Proteus polysaccharides in colony expansion and virulence requires a knowledge of their structures. Including the Cmf-CPS structure of this report, three Proteus CPS structures have been described in the literature and are shown in Fig. 4. Although these three structures are quite different from one another, they have two general similarities. First, all three structures are acidic in that they all contain at least one uronosyl residue; the CPS from P. mirabilis ATCC49565 has a branching terminally linked ␣-D-glucuronosyl residue, the P. vulgaris CPS contains a 4-linked ␣-D-glucuronosyl residue, and the P. mirabilis WT19 CPS contains 3-linked ␣-D-mannuronosyl and branching terminally linked ␤-D-galacturonsyl residues. Second, all three structures have amino sugars; P. mirabilis ATCC59565 CPS contains both N-acetylglucosamine and N-acetylfucosamine, P. vulgaris CP2-96 CPS contains N-acetylgalactosamine, and P. mirabilis WT19 CPS contains both N-acetylglucosamine and N-acetylgalactosamine. Understanding the molecular basis by which these acidic CPSs facilitate Proteus swarming will require further investigation. Fig. 3: s ϭ strong, m ϭ medium, and w ϭ weak.